Publish qualified Nox Choice semantics and latest Decision comparison

DIAGNOSTICS.md

@@ -6,9 +6,8 @@ These axes remain separate from headline accuracy. Probability metrics use the s
 
 | Model | Valid / requested | Brier ↓ | NLL ↓ | ECE % ↓ | Coverage at ≤5% error % ↑ | AURC ↓ |
 |---|---:|---:|---:|---:|---:|---:|
-| Jev | 1046/1046 | 0.1912 | 0.5743 | 3.35 | 76.96 | 0.0346 |
 | Lux | 1046/1046 | 0.3058 | 0.6201 | 5.13 | 57.36 | 0.0639 |
-| Nox | 1046/1046 | 0.4350 | 0.9375 | 10.27 | 35.18 | 0.1195 |
+| Nox | 1046/1046 | 0.4169 | 0.9079 | 11.80 | 36.42 | 0.1109 |
 | Kev-9B | 1046/1046 | 0.2948 | 0.6028 | 4.45 | 58.03 | 0.0586 |
 | Kev-4B | 1046/1046 | 0.3145 | 0.6434 | 2.61 | 56.79 | 0.0659 |
 | Qwen3.5-9B | 1046/1046 | 0.3673 | 0.7597 | 9.75 | 47.42 | 0.0872 |
@@ -19,6 +18,7 @@ These axes remain separate from headline accuracy. Probability metrics use the s
 | Qwen3.5-2B | 1046/1046 | 0.5708 | 1.2327 | 12.30 | 11.76 | 0.2699 |
 | Laya · English | 1046/1046 | 0.5832 | 1.2275 | 12.90 | 0.48 | 0.2780 |
 | Laya · Multilingual | 1046/1046 | 0.6873 | 1.4653 | 21.36 | 0.00 | 0.3460 |
+| Jev | 1046/1046 | 0.1912 | 0.5743 | 3.35 | 76.96 | 0.0346 |
 
 Coverage at an error threshold keeps whole confidence-tie groups together. These are empirical observed-sample results, not a deployment error guarantee.
 
@@ -26,9 +26,8 @@ Coverage at an error threshold keeps whole confidence-tie groups together. These
 
 | Model | Valid / requested pairs | Both correct % ↑ | Semantic flip % ↓ | Mean half-L1 ↓ |
 |---|---:|---:|---:|---:|
-| Jev | 36/36 | 86.11 | 0.00 | 0.0208 |
 | Lux | 36/36 | 77.78 | 8.33 | 0.0923 |
-| Nox | 36/36 | 58.33 | 25.00 | 0.1049 |
+| Nox | 36/36 | 63.89 | 16.67 | 0.0795 |
 | Kev-9B | 36/36 | 80.56 | 2.78 | 0.0615 |
 | Kev-4B | 36/36 | 77.78 | 5.56 | 0.0667 |
 | Qwen3.5-9B | 36/36 | 75.00 | 11.11 | 0.1157 |
@@ -39,6 +38,7 @@ Coverage at an error threshold keeps whole confidence-tie groups together. These
 | Qwen3.5-2B | 36/36 | 55.56 | 30.56 | 0.2163 |
 | Laya · English | 36/36 | 44.44 | 19.44 | 0.0928 |
 | Laya · Multilingual | 36/36 | 50.00 | 22.22 | 0.1491 |
+| Jev | 36/36 | 86.11 | 0.00 | 0.0208 |
 
 The 36 paired permutations test the same semantics under changed option order. Stable answers can still be wrong; both-correct rate therefore accompanies flip rate. Refused pairs remain in the both-correct denominator.
 
@@ -46,7 +46,6 @@ The 36 paired permutations test the same semantics under changed option order. S
 
 | Model | Valid / requested | Intact/control accuracy % ↑ | Mean max P % ↓ | P≥0.9 share % ↓ | Normalized entropy ↑ | Paired confidence drop pp ↑ |
 |---|---:|---:|---:|---:|---:|---:|
-| Jev | 110/110 | 93.64 | 62.04 | 16.36 | 0.7657 | 30.30 |
 | Lux | 110/110 | 84.55 | 68.81 | 19.09 | 0.6785 | 19.59 |
 | Nox | 110/110 | 72.73 | 78.65 | 27.27 | 0.5072 | 12.64 |
 | Kev-9B | 110/110 | 91.82 | 39.61 | 0.00 | 0.9981 | 53.19 |
@@ -59,6 +58,7 @@ The 36 paired permutations test the same semantics under changed option order. S
 | Qwen3.5-2B | 110/110 | 51.82 | 65.19 | 9.09 | 0.7370 | 4.68 |
 | Laya · English | 110/110 | 49.09 | 67.96 | 0.00 | 0.7497 | -5.58 |
 | Laya · Multilingual | 110/110 | 37.27 | 75.70 | 29.09 | 0.5307 | -0.49 |
+| Jev | 110/110 | 93.64 | 62.04 | 16.36 | 0.7657 | 30.30 |
 
 The 110 unknowable examples have no scored true class and are excluded from accuracy. Confidence is compared with matched evidence-bearing controls. Lower unsupported confidence and a positive evidence-removal confidence drop are desirable; these are not correctness scores.
 
@@ -66,7 +66,6 @@ The 110 unknowable examples have no scored true class and are excluded from accu
 
 | Model | Original probability rows | Transfer probability rows | Transfer truncated questions |
 |---|---:|---:|---:|
-| Jev | 2720/2720 | 1264/1264 | — / not observable |
 | Lux | 2720/2720 | 1264/1264 | 0 |
 | Nox | 2720/2720 | 1264/1264 | 0 |
 | Kev-9B | 2720/2720 | 1264/1264 | 0 |
@@ -79,6 +78,7 @@ The 110 unknowable examples have no scored true class and are excluded from accu
 | Qwen3.5-2B | 2720/2720 | 1264/1264 | 0 |
 | Laya · English | 2720/2720 | 1264/1264 | 34 |
 | Laya · Multilingual | 2720/2720 | 1264/1264 | 14 |
+| Jev | 2720/2720 | 1264/1264 | — / not observable |
 
 Transfer coverage includes all 1,264 questions: clean, unknown-evidence and order variants. Accuracy uses 1,046 clean knowable questions. Native refusals count as incorrect when accuracy applies, and absent probabilities are never invented. Laya retains upstream truncation. Server-side truncation for Jev cannot be observed.
 
@@ -86,9 +86,8 @@ Transfer coverage includes all 1,264 questions: clean, unknown-evidence and orde
 
 | Model | Overall % | 95% component-bootstrap interval |
 |---|---:|---:|
-| Jev | 81.05 | 79.70–82.35 |
 | Lux | 76.72 | 75.35–78.07 |
-| Nox | 72.84 | 71.33–74.31 |
+| Nox | 73.09 | 71.57–74.56 |
 | Kev-9B | 71.89 | 70.42–73.35 |
 | Kev-4B | 70.09 | 68.45–71.63 |
 | Qwen3.5-9B | 69.73 | 68.27–71.20 |
@@ -99,5 +98,6 @@ Transfer coverage includes all 1,264 questions: clean, unknown-evidence and orde
 | Qwen3.5-2B | 57.24 | 55.74–58.76 |
 | Laya · English | 51.03 | 49.43–52.68 |
 | Laya · Multilingual | 47.19 | 45.58–48.82 |
+| Jev | 81.05 | 79.70–82.35 |
 
 Intervals use 10,000 paired whole-source-component bootstrap draws. Related source, variant, parent, control and exact-state records stay together; all models receive the same draws. Five panels are sampled independently. These descriptive intervals do not create a positive-confidence-interval release requirement.

EVALUATION.md

@@ -4,9 +4,8 @@ The comparison covers **3,766 scored decisions across 54 tasks** and all 13 disp
 
 | Model | Size | Decisions | Composition | Reading | Inference | Transfer | Overall |
 |---|---:|---:|---:|---:|---:|---:|---:|
-| Jev | — | 79.10 | 66.38 | 94.53 | 89.79 | 87.19 | 81.05 |
+| Nox | 4B | **83.00** | **51.79** | 79.06 | **86.25** | 69.60 | **73.09** |
 | Lux | 9B | **83.21** | **51.88** | 90.31 | **90.83** | 77.44 | **76.72** |
-| Nox | 4B | **83.00** | **51.79** | 79.06 | **86.25** | 67.97 | **72.84** |
 | Kev-9B | 9B | 76.75 | 45.75 | 86.72 | 83.54 | 79.25 | 71.89 |
 | Kev-4B | 4B | 71.90 | 48.54 | 81.88 | 84.58 | 76.10 | 70.09 |
 | Qwen3.5-9B | 9B | 73.91 | 44.62 | 89.84 | 79.58 | 73.23 | 69.73 |
@@ -17,8 +16,9 @@ The comparison covers **3,766 scored decisions across 54 tasks** and all 13 disp
 | Qwen3.5-2B | 2B | 57.12 | 39.00 | 73.75 | 72.29 | 56.31 | 57.24 |
 | Laya · English | 0.421B | 56.54 | 35.33 | 51.41 | 63.75 | 53.06 | 51.03 |
 | Laya · Multilingual | 0.322B | 47.25 | 38.92 | 50.78 | 57.29 | 47.13 | 47.19 |
+| Jev | — | 79.10 | 66.38 | 94.53 | 89.79 | 87.19 | 81.05 |
 
-Accuracy (%). Rows are sorted by overall score. Bold marks Decision-family cells strictly above every external open or untuned reference in that metric; Jev and the other Decision models are excluded from that threshold.
+Accuracy (%). The current model is first, other open models follow by overall score, and Jev is the frontier reference at the end. Bold marks Decision-family cells strictly above every external open or untuned reference in that metric; Jev and the other Decision models are excluded from that threshold.
 
 ## Scope and weighting
 

MATERIALS.json

@@ -1,30 +1,26 @@
 {
-  "scope": "13model public comparison; presentation-only roster amendment",
-  "public_models": 13,
-  "task_rows": 54,
-  "presentation_amendment_sha256": "f078ecb12e7f4f2baaab522d07af6695c18d9c4669a5f61623e85c93b910fd1f",
-  "weights_runtime_temperature_unchanged": true,
+  "scope": "Latest13model comparison,54tasks and separate diagnostics.",
+  "statistics_sha256": "b319a993ac569ea1fa8da027ccb2195868557e4efe232a5d06091db2884086be",
   "files": {
-    "DIAGNOSTICS.md": "f0e3a5040f9843bf1480f0383d2f01f7e9a601347d6b1545d68bf861d1f64007",
-    "EVALUATION.md": "0624b6aebb35e86e8456ed734027ae4590fbf24c91775e4e228fe89454780f14",
-    "QUESTION-SCALING.md": "898f3554434284ff506dbc63c34859750f07b33dcf71f5b6d4ec7a798a56eb28",
-    "README.md": "02678d253920bc54d7af864edbee273bb649a5268ede892a6602d6b023f66800",
-    "SENSITIVITY.md": "6240f01d42aa4364ca2091f1e2d25619c21db91035348457609d0d2f9a740048",
-    "TASKS.md": "ffd8e2da2bed770855e69d54c44479bd66bf27a319684954ad8b55f742706341",
-    "assets/architecture.png": "8fcfa10b3941ee4aec1f5fba20a25e382a7d5de83074071c9ef68e627996de40",
-    "assets/architecture.svg": "215d9af6f94cc24847fc1d23a0b2a287b36afdc1e10ea4f5c82895b2fe103002",
-    "assets/decision-matrix.pdf": "12f963382e810eb2adb8a9fbbc3b3b7e25e75e4c13d56a04db81419e0b565033",
-    "assets/decision-matrix.png": "29c1fc0ad52b53bfc37f76788a9a824d5b767c8ed38c2474b763c1cb941182c4",
-    "assets/decision-matrix.svg": "95e90ed75b9d942e212b621a2968be3c5387d0bf637ca681c885062c5de93456",
-    "assets/decision-question-scaling.pdf": "d756f60c142ecaf647e59ea5bb3a611a39d1572a6471cd49afb7ef602a23fb3b",
-    "assets/decision-question-scaling.png": "722e664974b28324693f5f4af6609330c278385ac118ed6f757b8fc0afec64ac",
-    "assets/decision-question-scaling.svg": "cc591eafa335b5d16edd306d6368679afd283a4b14d5f83fc77780a352b1d315",
-    "assets/decision-ranking.pdf": "2c5d77b3a2420531e25a35e74389cd74968840117f90ab9f0f79e3e245101a97",
-    "assets/decision-ranking.png": "a201fa8f34771901d05b3162ed33de26f38bdba6dfbcab4579b13de017cce40f",
-    "assets/decision-ranking.svg": "39be265fee1f5e6a0b051524fadc33786c0a923a28d771a774ea3c01ace460dc",
-    "assets/readout.png": "fb6b3d1cf149a1a3ccb584eb1c213df73c07bcd328732b4238928e2a2a4af085",
-    "metrics/benchmark.json": "40b1da632cb586a5122fa34f85ea654e2890d6619bbe707b49afe829b4e8d3ce",
-    "metrics/evaluation-provenance.json": "cab27fc814933826d4a280ab280877c4683524a5ea67773e4e134a00fcf8befa",
-    "metrics/question-scaling.json": "9407f7ebf86581cdea21e704cdbebb9ced0a485eb980925f48233ebea0f830dc"
+    "README.md": "41dfa047d33bdc3aa9feff55928972533ba803a30b332ca380f042268fa1e848",
+    "EVALUATION.md": "4159b7939a107e91b13e47345a192eea489f8bd870943e4b397e094b33bba905",
+    "TASKS.md": "2fb04bc9c17d805e2abd5e90520df25e79d85aa5296cd42074d4b156f54bf862",
+    "DIAGNOSTICS.md": "0b290793565cc5007e9ee7bc79625812c73202c7d8bfce79d8816a0e7a93a965",
+    "SENSITIVITY.md": "c3ea31420db4f2e5f17cc7197197a3ddd18abdad773ffb5bf1171b67136c3bc2",
+    "WEIGHTING.md": "c3ea31420db4f2e5f17cc7197197a3ddd18abdad773ffb5bf1171b67136c3bc2",
+    "metrics/benchmark.json": "178e2e5f58da45f9fbf378ca82b49d7878fc8e9c93eee195a5ef1e7ba0f4fc6b",
+    "metrics/evaluation-provenance.json": "a3718fcee42a8cfe0144b01e04346c36a09ac7958154e4561327753d87d5fe8c",
+    "assets/decision-ranking.svg": "ab7ca66a9202587dee593714e46be760cb12ac91b6459074bcbbe11c59107d9d",
+    "assets/decision-ranking.pdf": "a28d4b0e8649cb7bf2db452a11d93675631af795561726aa63432b05b5888049",
+    "assets/decision-ranking.png": "d252593501e8faeb9c5c8cc0a9ae9a96591e7aa8696cf5394cdfec9c0ddac3b7",
+    "assets/decision-matrix.svg": "d99102fafe842f2a6bdd4d455f915833d5e0e09e3a22e5211fbf86e5647247ab",
+    "assets/decision-matrix.pdf": "5e127c0cd1a6f83b772a3e66ac2e1f4ec19913ac4938371c6f190a23c50a29fe",
+    "assets/decision-matrix.png": "c775737aedeecadb0824092d37014a8e2cffa90af0dafcf10d584e0db823ecf8",
+    "code/decision_api.py": "6716bdabca3cd2f1aa447d42a6cf53ee62d7ea97984a01f16b4c09fac8aaf17e",
+    "bundle-manifest.json": "7d9b06bc25a75b1f131aabd280df0ef2777bf69db98574a300067ace40c61c9c",
+    "NULL_DESCRIPTION_RENDERING.json": "3b531cab60cba35648ab71fe4be7fd1af619f02ee337e3beed6296fdfc8b9ec8",
+    "model-card-example.json": "1c4fc87d543722e2c3b839cbebd30c5d1370fcaf2d897f97de13079a668ac93e",
+    "USAGE.md": "b4c2f44ce120149c7cdee3749283958f34ba9caf4b6492442818c1b6a748c0d3",
+    "RUNTIME-RELEASE.json": "325d7e037acab5f61dab4d56d817e7d7e2e873d9b5f0d0df587afd6c33364a46"
   }
 }

NULL_DESCRIPTION_RENDERING.json

@@ -0,0 +1,7 @@
+{
+  "rule": "For Choice only, replace a null description with its original key text before tokenization.",
+  "preserved": "Non-null descriptions including empty string; keys/order; Noul and Score; original request; weights/tokenizer/temperature/numeric runtime.",
+  "source_bundle_manifest_sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830",
+  "qualification": "Candidate only; full fixed benchmark and independent public loading proof required.",
+  "not_a_weight_training_update": true
+}

README.md

@@ -32,13 +32,12 @@ tags:
 
 ## Measured capability
 
-**72.84% overall accuracy** across 3,766 scored decisions and 54 tasks. Nox leads Kev-4B by **2.75 percentage points** on this decision-focused comparison; reading and transfer remain opportunities to improve.
+**73.09% overall accuracy** across 3,766 scored decisions and 54 tasks. Nox leads Kev-4B by **2.99 percentage points** on this decision-focused comparison; reading and transfer remain opportunities to improve.
 
 | Model | Size | Decisions | Composition | Reading | Inference | Transfer | Overall |
 |---|---:|---:|---:|---:|---:|---:|---:|
-| Jev | — | 79.10 | 66.38 | 94.53 | 89.79 | 87.19 | 81.05 |
+| Nox | 4B | **83.00** | **51.79** | 79.06 | **86.25** | 69.60 | **73.09** |
 | Lux | 9B | **83.21** | **51.88** | 90.31 | **90.83** | 77.44 | **76.72** |
-| Nox | 4B | **83.00** | **51.79** | 79.06 | **86.25** | 67.97 | **72.84** |
 | Kev-9B | 9B | 76.75 | 45.75 | 86.72 | 83.54 | 79.25 | 71.89 |
 | Kev-4B | 4B | 71.90 | 48.54 | 81.88 | 84.58 | 76.10 | 70.09 |
 | Qwen3.5-9B | 9B | 73.91 | 44.62 | 89.84 | 79.58 | 73.23 | 69.73 |
@@ -49,6 +48,7 @@ tags:
 | Qwen3.5-2B | 2B | 57.12 | 39.00 | 73.75 | 72.29 | 56.31 | 57.24 |
 | Laya · English | 0.421B | 56.54 | 35.33 | 51.41 | 63.75 | 53.06 | 51.03 |
 | Laya · Multilingual | 0.322B | 47.25 | 38.92 | 50.78 | 57.29 | 47.13 | 47.19 |
+| Jev | — | 79.10 | 66.38 | 94.53 | 89.79 | 87.19 | 81.05 |
 
 Accuracy (%). Overall weights: Decisions **30%**, Composition **25%**, Reading **15%**, Inference **15%**, Transfer **15%**. These outcome-informed product-priority weights were chosen after observing results; reweighting is not a training improvement. Bold marks Decision-family cells above every external open or untuned reference for that metric, excluding Jev and the other Decision models.
 
@@ -78,6 +78,8 @@ print(model.decide(**REQUEST)["answers"])
 
 [Tested SystemOne request and output](model-card-example.json) · [Install and typed API guide](USAGE.md)
 
+Choice candidates with a null description use their ID text, which may increase input tokens.
+
 The complete state, question and candidates must fit 16,384 tokens; overflow is rejected. The bundled normalization profile loads automatically. AMD gfx942 is validated; CPU/MPS are unsupported and NVIDIA is unqualified. Use a fresh Python process when switching profiles.
 
 ## Architecture

RUNTIME-RELEASE.json

@@ -1,30 +1,11 @@
 {
-  "format": "decision-runtime-patch-v1",
-  "family": "Nox",
-  "release_tag": "v1.3.1",
-  "change_kind": "runtime_only",
-  "weights_revision": "ad089ad3a5dc9a7a21e6d96db546bb53e2212654",
-  "source_bundle_manifest_sha256": "92d7f5be5e1ef21ee682f574b35cc01de4dd8ab16b8aa6edf0dfbdfcfcfabba3",
-  "bundle_manifest_sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830",
-  "changed_inference_files": [
-    "code/decision_api.py"
-  ],
-  "weights_tokenizer_prompts_calibration_profile_unchanged": true,
-  "default_public_parity": {
-    "requests": 2856,
-    "answers": 3160,
-    "fixtures": 58,
-    "raw_logits_probabilities_typed_outputs_exact": true,
-    "receipt_sha256": "946dba5c31074b63efa3223ede17857e5f9d1b7b3b44de08cd9add9ceece878c"
-  },
-  "timing": {
-    "receipt_sha256": "501f25efd1813b566d77e98278229440b7a3b4a2358def1462c0171a33e9f3be",
-    "shape": "distinct fixed-length questions; Q=32; 499 tokens per question",
-    "reference_ms": 419.124,
-    "candidate_ms": 414.199,
-    "same_measured_api_sha256": "273f6f10f22d5a68b8db34cfcbd35407fb43d8030f6d7cd188bcf118dd90a152",
-    "no_shared_state_neural_cache_claim": true
-  },
-  "capability_scores_unchanged": true,
-  "downloaded_package_offline_proof_required_before_promotion": true
+  "release": "v1.3.2",
+  "kind": "SystemOne Choice null-description semantics",
+  "rule": "A null Choice description uses the original candidate key as its description.",
+  "weights_tokenizer_temperature_unchanged": true,
+  "explicit_descriptions_and_Noul_Score_unchanged": true,
+  "bundle_manifest_sha256": "7d9b06bc25a75b1f131aabd280df0ef2777bf69db98574a300067ace40c61c9c",
+  "qualified_statistics_sha256": "b319a993ac569ea1fa8da027ccb2195868557e4efe232a5d06091db2884086be",
+  "exact_public_contract_proof_sha256": "ebf27c7aa2bb4a12ba2cbb27e4096e46cb43ac4a20b9d3e1bc34f62777a1e63a",
+  "runtime_note": "This is an API rendering improvement, not a newly trained checkpoint."
 }

SENSITIVITY.md

@@ -1,10 +1,9 @@
-Product weights were changed after earlier results were observed. These are identical model predictions under different weights, not trained-model improvements.
+The same latest model predictions are shown under three weighting schemes. Product weights were chosen after earlier results were observed; this sensitivity view is not a trained-model improvement.
 
 | Model | Current 30/25/15/15/15 | Earlier 25/25/15/15/20 | Original four-panel mean |
 |---|---:|---:|---:|
-| Jev | 81.05 | 81.45 | 82.45 |
 | Lux | 76.72 | 76.43 | 79.06 |
-| Nox | 72.84 | 72.09 | 75.03 |
+| Nox | 73.09 | 72.42 | 75.03 |
 | Kev-9B | 71.89 | 72.01 | 73.19 |
 | Kev-4B | 70.09 | 70.30 | 71.73 |
 | Qwen3.5-9B | 69.73 | 69.70 | 71.99 |
@@ -15,3 +14,4 @@ Product weights were changed after earlier results were observed. These are iden
 | Qwen3.5-2B | 57.24 | 57.20 | 60.54 |
 | Laya · English | 51.03 | 50.85 | 51.76 |
 | Laya · Multilingual | 47.19 | 47.18 | 48.56 |
+| Jev | 81.05 | 81.45 | 82.45 |

TASKS.md

@@ -5,93 +5,93 @@ Accuracy (%) on the same requested rows. Bold marks a Decision-family result str
 <details>
 <summary>Decisions · 10 tasks</summary>
 
-| Task | n | Jev | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| News classification | 128 | 85.16 | 88.28 | 85.16 | 87.50 | 87.50 | 85.94 | 86.72 | 84.38 | 83.59 | 86.72 | 80.47 | 91.41 | 89.84 |
-| Boolean constraints | 64 | 100.00 | 95.31 | 93.75 | 95.31 | 71.88 | 98.44 | 71.88 | 62.50 | 50.00 | 81.25 | 37.50 | 43.75 | 39.06 |
-| Entity classification | 112 | 96.43 | 96.43 | 96.43 | 98.21 | 98.21 | 96.43 | 98.21 | 97.32 | 95.54 | 96.43 | 92.86 | 83.93 | 57.14 |
-| Intent routing | 64 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 96.88 | 81.25 | 81.25 |
-| Evidence placement | 96 | 30.21 | 61.46 | **100.00** | 50.00 | 41.67 | 79.17 | 8.33 | 72.92 | **98.96** | 33.33 | 9.38 | 95.83 | 54.17 |
-| Ordered rubric | 64 | 100.00 | 100.00 | 89.06 | 96.88 | 100.00 | 84.38 | 84.38 | 90.62 | 65.62 | 59.38 | 81.25 | 18.75 | 12.50 |
-| Relation composition | 96 | 56.25 | 61.46 | 52.08 | 51.04 | 62.50 | 36.46 | 54.17 | 51.04 | 37.50 | 44.79 | 48.96 | 25.00 | 35.42 |
-| Scoped evidence | 96 | 89.58 | **98.96** | **78.12** | 65.62 | 47.92 | 55.21 | 47.92 | 51.04 | **79.17** | 10.42 | 36.46 | 37.50 | 36.46 |
-| State tracking | 96 | 33.33 | 33.33 | 35.42 | 38.54 | 34.38 | 31.25 | 29.17 | 28.12 | 27.08 | 31.25 | 25.00 | 23.96 | 29.17 |
-| In / out of menu | 64 | 100.00 | **96.88** | **100.00** | 84.38 | 75.00 | 71.88 | 59.38 | 60.94 | **100.00** | 57.81 | 62.50 | 64.06 | 37.50 |
+| Task | n | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual | Jev |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| News classification | 128 | 88.28 | 85.16 | 87.50 | 87.50 | 85.94 | 86.72 | 84.38 | 83.59 | 86.72 | 80.47 | 91.41 | 89.84 | 85.16 |
+| Boolean constraints | 64 | 95.31 | 93.75 | 95.31 | 71.88 | 98.44 | 71.88 | 62.50 | 50.00 | 81.25 | 37.50 | 43.75 | 39.06 | 100.00 |
+| Entity classification | 112 | 96.43 | 96.43 | 98.21 | 98.21 | 96.43 | 98.21 | 97.32 | 95.54 | 96.43 | 92.86 | 83.93 | 57.14 | 96.43 |
+| Intent routing | 64 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 96.88 | 81.25 | 81.25 | 100.00 |
+| Evidence placement | 96 | 61.46 | **100.00** | 50.00 | 41.67 | 79.17 | 8.33 | 72.92 | **98.96** | 33.33 | 9.38 | 95.83 | 54.17 | 30.21 |
+| Ordered rubric | 64 | 100.00 | 89.06 | 96.88 | 100.00 | 84.38 | 84.38 | 90.62 | 65.62 | 59.38 | 81.25 | 18.75 | 12.50 | 100.00 |
+| Relation composition | 96 | 61.46 | 52.08 | 51.04 | 62.50 | 36.46 | 54.17 | 51.04 | 37.50 | 44.79 | 48.96 | 25.00 | 35.42 | 56.25 |
+| Scoped evidence | 96 | **98.96** | **78.12** | 65.62 | 47.92 | 55.21 | 47.92 | 51.04 | **79.17** | 10.42 | 36.46 | 37.50 | 36.46 | 89.58 |
+| State tracking | 96 | 33.33 | 35.42 | 38.54 | 34.38 | 31.25 | 29.17 | 28.12 | 27.08 | 31.25 | 25.00 | 23.96 | 29.17 | 33.33 |
+| In / out of menu | 64 | **96.88** | **100.00** | 84.38 | 75.00 | 71.88 | 59.38 | 60.94 | **100.00** | 57.81 | 62.50 | 64.06 | 37.50 | 100.00 |
 
 </details>
 
 <details>
 <summary>Composition · 10 tasks</summary>
 
-| Task | n | Jev | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| Record identity | 80 | 76.25 | **71.25** | 53.75 | 45.00 | 50.00 | 57.50 | 56.25 | 50.00 | 50.00 | 50.00 | 46.25 | 47.50 | 50.00 |
-| Capacity assignment | 80 | 68.75 | **71.25** | 46.25 | 50.00 | 50.00 | 50.00 | 51.25 | 50.00 | 47.50 | 50.00 | 50.00 | 63.75 | 50.00 |
-| Constraint assignment | 80 | 55.00 | **43.75** | **41.25** | 27.50 | 32.50 | 25.00 | 23.75 | 23.75 | 26.25 | 20.00 | 21.25 | 22.50 | 27.50 |
-| Intent routing · EN | 120 | 91.67 | 90.83 | 91.67 | 86.67 | 87.50 | 88.33 | 90.00 | 91.67 | 90.83 | 83.33 | 70.83 | 74.17 | 70.83 |
-| Intent routing · ZH | 120 | 88.33 | 86.67 | 87.50 | 85.83 | 86.67 | 86.67 | 88.33 | 86.67 | 87.50 | 85.83 | 74.17 | 44.17 | 73.33 |
-| Multiset reconciliation | 80 | 58.75 | 26.25 | 28.75 | 37.50 | 28.75 | 20.00 | 23.75 | 27.50 | 28.75 | 25.00 | 27.50 | 26.25 | 32.50 |
-| Ordered service loss | 80 | 42.50 | 20.00 | **30.00** | 25.00 | 20.00 | 20.00 | 22.50 | 21.25 | 25.00 | 27.50 | 20.00 | 20.00 | 17.50 |
-| Conflicting rule closure | 80 | 66.25 | **36.25** | 33.75 | 27.50 | 33.75 | 27.50 | 26.25 | 25.00 | 25.00 | 27.50 | 27.50 | 18.75 | 20.00 |
-| Temporal exclusion | 80 | 41.25 | 21.25 | 42.50 | 28.75 | 45.00 | 36.25 | 42.50 | 31.25 | 41.25 | 21.25 | 28.75 | 12.50 | 28.75 |
-| Transaction recovery | 80 | 75.00 | 51.25 | **62.50** | 43.75 | 51.25 | 35.00 | 41.25 | 26.25 | 38.75 | 32.50 | 23.75 | 23.75 | 18.75 |
+| Task | n | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual | Jev |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| Record identity | 80 | **71.25** | 53.75 | 45.00 | 50.00 | 57.50 | 56.25 | 50.00 | 50.00 | 50.00 | 46.25 | 47.50 | 50.00 | 76.25 |
+| Capacity assignment | 80 | **71.25** | 46.25 | 50.00 | 50.00 | 50.00 | 51.25 | 50.00 | 47.50 | 50.00 | 50.00 | 63.75 | 50.00 | 68.75 |
+| Constraint assignment | 80 | **43.75** | **41.25** | 27.50 | 32.50 | 25.00 | 23.75 | 23.75 | 26.25 | 20.00 | 21.25 | 22.50 | 27.50 | 55.00 |
+| Intent routing · EN | 120 | 90.83 | 91.67 | 86.67 | 87.50 | 88.33 | 90.00 | 91.67 | 90.83 | 83.33 | 70.83 | 74.17 | 70.83 | 91.67 |
+| Intent routing · ZH | 120 | 86.67 | 87.50 | 85.83 | 86.67 | 86.67 | 88.33 | 86.67 | 87.50 | 85.83 | 74.17 | 44.17 | 73.33 | 88.33 |
+| Multiset reconciliation | 80 | 26.25 | 28.75 | 37.50 | 28.75 | 20.00 | 23.75 | 27.50 | 28.75 | 25.00 | 27.50 | 26.25 | 32.50 | 58.75 |
+| Ordered service loss | 80 | 20.00 | **30.00** | 25.00 | 20.00 | 20.00 | 22.50 | 21.25 | 25.00 | 27.50 | 20.00 | 20.00 | 17.50 | 42.50 |
+| Conflicting rule closure | 80 | **36.25** | 33.75 | 27.50 | 33.75 | 27.50 | 26.25 | 25.00 | 25.00 | 27.50 | 27.50 | 18.75 | 20.00 | 66.25 |
+| Temporal exclusion | 80 | 21.25 | 42.50 | 28.75 | 45.00 | 36.25 | 42.50 | 31.25 | 41.25 | 21.25 | 28.75 | 12.50 | 28.75 | 41.25 |
+| Transaction recovery | 80 | 51.25 | **62.50** | 43.75 | 51.25 | 35.00 | 41.25 | 26.25 | 38.75 | 32.50 | 23.75 | 23.75 | 18.75 | 75.00 |
 
 </details>
 
 <details>
 <summary>Reading · 3 tasks</summary>
 
-| Task | n | Jev | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| Yes / no reading | 160 | 92.50 | 90.62 | 86.25 | 91.25 | 88.75 | 84.38 | 91.88 | 83.75 | 83.12 | 74.38 | 65.00 | 69.38 | 69.38 |
-| Reading · EN | 160 | 96.88 | 91.25 | 73.12 | 83.12 | 75.62 | 98.75 | 93.12 | 93.12 | 70.00 | 63.75 | 83.75 | 38.12 | 36.25 |
-| Reading · ZH | 160 | 96.25 | 88.75 | 70.62 | 81.25 | 74.38 | 91.88 | 91.25 | 91.25 | 70.00 | 58.75 | 81.25 | 28.75 | 28.12 |
+| Task | n | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual | Jev |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| Yes / no reading | 160 | 90.62 | 86.25 | 91.25 | 88.75 | 84.38 | 91.88 | 83.75 | 83.12 | 74.38 | 65.00 | 69.38 | 69.38 | 92.50 |
+| Reading · EN | 160 | 91.25 | 73.12 | 83.12 | 75.62 | 98.75 | 93.12 | 93.12 | 70.00 | 63.75 | 83.75 | 38.12 | 36.25 | 96.88 |
+| Reading · ZH | 160 | 88.75 | 70.62 | 81.25 | 74.38 | 91.88 | 91.25 | 91.25 | 70.00 | 58.75 | 81.25 | 28.75 | 28.12 | 96.25 |
 
 </details>
 
 <details>
 <summary>Inference · 4 tasks</summary>
 
-| Task | n | Jev | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| Contextual reasoning | 120 | 86.67 | **83.33** | 70.83 | 68.33 | 70.83 | 64.17 | 66.67 | 60.83 | 69.17 | 42.50 | 56.67 | 30.00 | 20.00 |
-| Answerability | 120 | 90.83 | **93.33** | **88.33** | 78.33 | 81.67 | 74.17 | 84.17 | 81.67 | 83.33 | 66.67 | 82.50 | 57.50 | 55.00 |
-| Textual entailment | 120 | 82.50 | 90.00 | 89.17 | 90.83 | 89.17 | 81.67 | 90.83 | 80.00 | 88.33 | 77.50 | 64.17 | 72.50 | 72.50 |
-| Scientific inference | 120 | 99.17 | 96.67 | 96.67 | 96.67 | 96.67 | 98.33 | 95.83 | 96.67 | 95.83 | 88.33 | 85.83 | 95.00 | 81.67 |
+| Task | n | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual | Jev |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| Contextual reasoning | 120 | **83.33** | 70.83 | 68.33 | 70.83 | 64.17 | 66.67 | 60.83 | 69.17 | 42.50 | 56.67 | 30.00 | 20.00 | 86.67 |
+| Answerability | 120 | **93.33** | **88.33** | 78.33 | 81.67 | 74.17 | 84.17 | 81.67 | 83.33 | 66.67 | 82.50 | 57.50 | 55.00 | 90.83 |
+| Textual entailment | 120 | 90.00 | 89.17 | 90.83 | 89.17 | 81.67 | 90.83 | 80.00 | 88.33 | 77.50 | 64.17 | 72.50 | 72.50 | 82.50 |
+| Scientific inference | 120 | 96.67 | 96.67 | 96.67 | 96.67 | 98.33 | 95.83 | 96.67 | 95.83 | 88.33 | 85.83 | 95.00 | 81.67 | 99.17 |
 
 </details>
 
 <details>
 <summary>Transfer · 27 tasks</summary>
 
-| Task | n | Jev | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| Buried emotion | 20 | 70.00 | 40.00 | 20.00 | 65.00 | 55.00 | 40.00 | 85.00 | 45.00 | 20.00 | 40.00 | 55.00 | 60.00 | 35.00 |
-| Buried paraphrase | 20 | 95.00 | **85.00** | 75.00 | 80.00 | 75.00 | 80.00 | 75.00 | 75.00 | 70.00 | 60.00 | 70.00 | 55.00 | 70.00 |
-| Buried entailment | 20 | 95.00 | 90.00 | 90.00 | 95.00 | 90.00 | 95.00 | 90.00 | 75.00 | 75.00 | 85.00 | 65.00 | 60.00 | 55.00 |
-| Buried offensive-language detection | 20 | 55.00 | 35.00 | 45.00 | 80.00 | 65.00 | 45.00 | 80.00 | 40.00 | 35.00 | 60.00 | 35.00 | 70.00 | 80.00 |
-| Combined policy conditions | 32 | 96.88 | **90.62** | 75.00 | 68.75 | 78.12 | 53.12 | 53.12 | 59.38 | 78.12 | 40.62 | 62.50 | 62.50 | 43.75 |
-| Policy exceptions | 32 | 100.00 | 96.88 | 96.88 | 87.50 | 96.88 | 59.38 | 56.25 | 75.00 | 71.88 | 65.62 | 40.62 | 43.75 | 37.50 |
-| Policy negation | 32 | 90.62 | 87.50 | 87.50 | 90.62 | 87.50 | 62.50 | 62.50 | 53.12 | 71.88 | 71.88 | 46.88 | 53.12 | 50.00 |
-| Authorization contrast | 40 | 100.00 | 100.00 | 97.50 | 100.00 | 100.00 | 100.00 | 100.00 | 97.50 | 50.00 | 97.50 | 62.50 | 67.50 | 50.00 |
-| Deadline contrast | 40 | 92.50 | 87.50 | 70.00 | 87.50 | 75.00 | 77.50 | 30.00 | 52.50 | 40.00 | 45.00 | 25.00 | 35.00 | 22.50 |
-| Emotion | 80 | 67.50 | 61.25 | 40.00 | 67.50 | 65.00 | 52.50 | 86.25 | 65.00 | 22.50 | 60.00 | 67.50 | 62.50 | 56.25 |
-| MMLU | 80 | 88.75 | **77.50** | 61.25 | 75.00 | 68.75 | 73.75 | 62.50 | 66.25 | 52.50 | 51.25 | 53.75 | 22.50 | 27.50 |
-| MMLU-Pro | 200 | 84.00 | 53.00 | 37.00 | 53.00 | 45.50 | 53.50 | 37.50 | 45.00 | 24.50 | 22.50 | 28.00 | 11.00 | 11.50 |
-| Paraphrase | 80 | 87.50 | **91.25** | 85.00 | 81.25 | 81.25 | 88.75 | 77.50 | 83.75 | 80.00 | 58.75 | 70.00 | 86.25 | 75.00 |
-| Question entailment | 80 | 91.25 | 93.75 | 91.25 | 95.00 | 91.25 | 90.00 | 87.50 | 87.50 | 83.75 | 81.25 | 63.75 | 80.00 | 73.75 |
-| Science questions | 80 | 100.00 | 100.00 | 98.75 | 100.00 | 100.00 | 100.00 | 98.75 | 98.75 | 97.50 | 96.25 | 96.25 | 90.00 | 72.50 |
-| Offensive-language detection | 80 | 76.25 | 73.75 | 75.00 | 86.25 | 85.00 | 83.75 | 88.75 | 77.50 | 75.00 | 72.50 | 83.75 | 81.25 | 82.50 |
-| Evidence control · age eligibility | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 90.00 | 50.00 |
-| Evidence control · authorization | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 80.00 | 100.00 | 50.00 | 100.00 | 50.00 | 70.00 | 50.00 |
-| Evidence control · deadline | 10 | 100.00 | 80.00 | 60.00 | 70.00 | 60.00 | 80.00 | 30.00 | 50.00 | 30.00 | 40.00 | 30.00 | 30.00 | 30.00 |
-| Evidence control · late fee | 10 | 100.00 | 90.00 | 90.00 | 100.00 | 100.00 | 90.00 | 70.00 | 80.00 | 80.00 | 90.00 | 70.00 | 40.00 | 40.00 |
-| Evidence control · quantity limit | 10 | 100.00 | 100.00 | 80.00 | 100.00 | 100.00 | 70.00 | 80.00 | 70.00 | 70.00 | 100.00 | 20.00 | 60.00 | 30.00 |
-| Evidence control · return window | 10 | 50.00 | 60.00 | 40.00 | 70.00 | 70.00 | 60.00 | 60.00 | 40.00 | 40.00 | 80.00 | 50.00 | 60.00 | 40.00 |
-| Evidence control · shipping delay | 10 | 90.00 | 90.00 | 40.00 | 90.00 | 80.00 | 80.00 | 80.00 | 60.00 | 80.00 | 100.00 | 30.00 | 30.00 | 30.00 |
-| Evidence control · sla response | 10 | 100.00 | 70.00 | 40.00 | 100.00 | 100.00 | 50.00 | 30.00 | 50.00 | 40.00 | 100.00 | 50.00 | 50.00 | 30.00 |
-| Evidence control · spend threshold | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 90.00 | 100.00 | 100.00 | 50.00 | 50.00 |
-| Evidence control · volume discount | 10 | 100.00 | 100.00 | 80.00 | 100.00 | 100.00 | 100.00 | 90.00 | 100.00 | 90.00 | 100.00 | 20.00 | 20.00 | 30.00 |
-| Evidence control · warranty claim | 10 | 90.00 | 40.00 | 70.00 | 80.00 | 100.00 | 60.00 | 40.00 | 50.00 | 50.00 | 50.00 | 50.00 | 40.00 | 30.00 |
+| Task | n | Lux | Nox | Kev-9B | Kev-4B | Qwen3.5-9B | Decider | Qwen3.5-4B | Sol | Kev-0.8B | Qwen3.5-2B | Laya · English | Laya · Multilingual | Jev |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| Buried emotion | 20 | 40.00 | 25.00 | 65.00 | 55.00 | 40.00 | 85.00 | 45.00 | 20.00 | 40.00 | 55.00 | 60.00 | 35.00 | 70.00 |
+| Buried paraphrase | 20 | **85.00** | 75.00 | 80.00 | 75.00 | 80.00 | 75.00 | 75.00 | 70.00 | 60.00 | 70.00 | 55.00 | 70.00 | 95.00 |
+| Buried entailment | 20 | 90.00 | 90.00 | 95.00 | 90.00 | 95.00 | 90.00 | 75.00 | 75.00 | 85.00 | 65.00 | 60.00 | 55.00 | 95.00 |
+| Buried offensive-language detection | 20 | 35.00 | 45.00 | 80.00 | 65.00 | 45.00 | 80.00 | 40.00 | 35.00 | 60.00 | 35.00 | 70.00 | 80.00 | 55.00 |
+| Combined policy conditions | 32 | **90.62** | 75.00 | 68.75 | 78.12 | 53.12 | 53.12 | 59.38 | 78.12 | 40.62 | 62.50 | 62.50 | 43.75 | 96.88 |
+| Policy exceptions | 32 | 96.88 | 96.88 | 87.50 | 96.88 | 59.38 | 56.25 | 75.00 | 71.88 | 65.62 | 40.62 | 43.75 | 37.50 | 100.00 |
+| Policy negation | 32 | 87.50 | 87.50 | 90.62 | 87.50 | 62.50 | 62.50 | 53.12 | 71.88 | 71.88 | 46.88 | 53.12 | 50.00 | 90.62 |
+| Authorization contrast | 40 | 100.00 | 97.50 | 100.00 | 100.00 | 100.00 | 100.00 | 97.50 | 50.00 | 97.50 | 62.50 | 67.50 | 50.00 | 100.00 |
+| Deadline contrast | 40 | 87.50 | 70.00 | 87.50 | 75.00 | 77.50 | 30.00 | 52.50 | 40.00 | 45.00 | 25.00 | 35.00 | 22.50 | 92.50 |
+| Emotion | 80 | 61.25 | 60.00 | 67.50 | 65.00 | 52.50 | 86.25 | 65.00 | 22.50 | 60.00 | 67.50 | 62.50 | 56.25 | 67.50 |
+| MMLU | 80 | **77.50** | 61.25 | 75.00 | 68.75 | 73.75 | 62.50 | 66.25 | 52.50 | 51.25 | 53.75 | 22.50 | 27.50 | 88.75 |
+| MMLU-Pro | 200 | 53.00 | 37.00 | 53.00 | 45.50 | 53.50 | 37.50 | 45.00 | 24.50 | 22.50 | 28.00 | 11.00 | 11.50 | 84.00 |
+| Paraphrase | 80 | **91.25** | 85.00 | 81.25 | 81.25 | 88.75 | 77.50 | 83.75 | 80.00 | 58.75 | 70.00 | 86.25 | 75.00 | 87.50 |
+| Question entailment | 80 | 93.75 | 91.25 | 95.00 | 91.25 | 90.00 | 87.50 | 87.50 | 83.75 | 81.25 | 63.75 | 80.00 | 73.75 | 91.25 |
+| Science questions | 80 | 100.00 | 98.75 | 100.00 | 100.00 | 100.00 | 98.75 | 98.75 | 97.50 | 96.25 | 96.25 | 90.00 | 72.50 | 100.00 |
+| Offensive-language detection | 80 | 73.75 | 75.00 | 86.25 | 85.00 | 83.75 | 88.75 | 77.50 | 75.00 | 72.50 | 83.75 | 81.25 | 82.50 | 76.25 |
+| Evidence control · age eligibility | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 90.00 | 50.00 | 100.00 |
+| Evidence control · authorization | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 80.00 | 100.00 | 50.00 | 100.00 | 50.00 | 70.00 | 50.00 | 100.00 |
+| Evidence control · deadline | 10 | 80.00 | 60.00 | 70.00 | 60.00 | 80.00 | 30.00 | 50.00 | 30.00 | 40.00 | 30.00 | 30.00 | 30.00 | 100.00 |
+| Evidence control · late fee | 10 | 90.00 | 90.00 | 100.00 | 100.00 | 90.00 | 70.00 | 80.00 | 80.00 | 90.00 | 70.00 | 40.00 | 40.00 | 100.00 |
+| Evidence control · quantity limit | 10 | 100.00 | 80.00 | 100.00 | 100.00 | 70.00 | 80.00 | 70.00 | 70.00 | 100.00 | 20.00 | 60.00 | 30.00 | 100.00 |
+| Evidence control · return window | 10 | 60.00 | 40.00 | 70.00 | 70.00 | 60.00 | 60.00 | 40.00 | 40.00 | 80.00 | 50.00 | 60.00 | 40.00 | 50.00 |
+| Evidence control · shipping delay | 10 | 90.00 | 40.00 | 90.00 | 80.00 | 80.00 | 80.00 | 60.00 | 80.00 | 100.00 | 30.00 | 30.00 | 30.00 | 90.00 |
+| Evidence control · sla response | 10 | 70.00 | 40.00 | 100.00 | 100.00 | 50.00 | 30.00 | 50.00 | 40.00 | 100.00 | 50.00 | 50.00 | 30.00 | 100.00 |
+| Evidence control · spend threshold | 10 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 | 90.00 | 100.00 | 100.00 | 50.00 | 50.00 | 100.00 |
+| Evidence control · volume discount | 10 | 100.00 | 80.00 | 100.00 | 100.00 | 100.00 | 90.00 | 100.00 | 90.00 | 100.00 | 20.00 | 20.00 | 30.00 | 100.00 |
+| Evidence control · warranty claim | 10 | 40.00 | 70.00 | 80.00 | 100.00 | 60.00 | 40.00 | 50.00 | 50.00 | 50.00 | 50.00 | 40.00 | 30.00 | 90.00 |
 
 </details>

USAGE.md

@@ -49,12 +49,14 @@ The model card must publish output from that model's real run; this document inv
 
 | Type | Input criteria | Answer |
 |---|---|---|
-| Choice | Ordered mapping of 2–255 external IDs to complete descriptions | `probabilities`, selected `choice` ID, `confidence` |
+| Choice | Ordered mapping of 2–255 external IDs to descriptions or null | `probabilities`, selected `choice` ID, `confidence` |
 | Noul | Optional mapping containing only `false` / `true` descriptions | `noul`: P(true); a hard judgment uses `>= 0.5` |
 | Score | Ordered list of 2–10 rubric descriptions | `probabilities` over string indices, expected level index `score`, `legend`, `confidence` |
 
 Response shape is `{"model": name, "answers": {question_name: answer}, "usage": {"input_tokens": total, "scored_questions": count}}`. Choice ties select the earliest candidate in insertion order. Score returns the expected ordinal **index**, not an arbitrary supplied numeric value; this adapter does not implement a supplied-values extension. Noul 0.5 is interpreted as true. Confidence is `(K * max(p) - 1)/(K - 1)`, clipped to [0,1], and is not a claimed reproduction of Jev's confidence statistic. Shipped temperature calibration does not make every confidence value a correctness guarantee.
 
+For Choice, a null description uses the original candidate ID as its semantic description. Explicit descriptions, including an empty string, are preserved. The original request object is not mutated. Noul and Score are unchanged.
+
 Question names are preserved as opaque bookkeeping IDs; candidate IDs and descriptions use the frozen renderer. Native objects are deterministically serialized with sorted JSON keys; strings preserve their contents. Do not interpret object/string field-order differences as identical token inputs. Non-finite or non-JSON inputs are rejected.
 
 The bound is **16,384 tokens per complete question**, including its state, instructions, all candidates and readout suffix. Questions are separate sequences, grouped in fixed batches of eight; the state is repeated for each question and counted repeatedly in `usage.input_tokens`. All questions are encoded before any forward pass. If any exceeds the bound, the whole call raises `ValueError` without truncation or partial answers. A lower `max_length` can be chosen at load time; a higher limit is rejected. This native wrapper does not impose the Studio's separate 16-question UI limit.

WEIGHTING.md

@@ -0,0 +1,17 @@
+The same latest model predictions are shown under three weighting schemes. Product weights were chosen after earlier results were observed; this sensitivity view is not a trained-model improvement.
+
+| Model | Current 30/25/15/15/15 | Earlier 25/25/15/15/20 | Original four-panel mean |
+|---|---:|---:|---:|
+| Lux | 76.72 | 76.43 | 79.06 |
+| Nox | 73.09 | 72.42 | 75.03 |
+| Kev-9B | 71.89 | 72.01 | 73.19 |
+| Kev-4B | 70.09 | 70.30 | 71.73 |
+| Qwen3.5-9B | 69.73 | 69.70 | 71.99 |
+| Decider | 67.71 | 67.97 | 71.75 |
+| Qwen3.5-4B | 67.29 | 67.24 | 70.25 |
+| Sol | 66.32 | 65.48 | 70.14 |
+| Kev-0.8B | 58.28 | 58.33 | 59.75 |
+| Qwen3.5-2B | 57.24 | 57.20 | 60.54 |
+| Laya · English | 51.03 | 50.85 | 51.76 |
+| Laya · Multilingual | 47.19 | 47.18 | 48.56 |
+| Jev | 81.05 | 81.45 | 82.45 |

bundle-manifest.json

@@ -1,12 +1,17 @@
 {
   "format": "research-pointer-bundle-v1",
-  "status": "joint-serving-candidate-awaiting-default-public-parity",
+  "status": "null-description-candidate-awaiting-public-proof-and-full-regression",
   "files": [
     {
       "file": "NORMALIZATION_RUNTIME.md",
       "bytes": 756,
       "sha256": "cf8e6ce1f07687a68b6adeb98e6704b8f292cb6adc7b10e4b84e8e70aa5472e5"
     },
+    {
+      "file": "NULL_DESCRIPTION_RENDERING.json",
+      "bytes": 514,
+      "sha256": "3b531cab60cba35648ab71fe4be7fd1af619f02ee337e3beed6296fdfc8b9ec8"
+    },
     {
       "file": "RUNTIME_BINDING.json",
       "bytes": 2621,
@@ -54,8 +59,8 @@
     },
     {
       "file": "code/decision_api.py",
-      "bytes": 10952,
-      "sha256": "273f6f10f22d5a68b8db34cfcbd35407fb43d8030f6d7cd188bcf118dd90a152"
+      "bytes": 10978,
+      "sha256": "6716bdabca3cd2f1aa447d42a6cf53ee62d7ea97984a01f16b4c09fac8aaf17e"
     },
     {
       "file": "code/decision_model.py",
@@ -221,7 +226,7 @@
     }
   ],
   "source_model_code_sha256": "d3e28489c09f3bd7130e2d43d92e0b5c4a08e25b09b21303904defb0ff1c3646",
-  "source_api_code_sha256": "273f6f10f22d5a68b8db34cfcbd35407fb43d8030f6d7cd188bcf118dd90a152",
+  "source_api_code_sha256": "6716bdabca3cd2f1aa447d42a6cf53ee62d7ea97984a01f16b4c09fac8aaf17e",
   "dev_sha256": "45b4cd46acc4be53b95c7ed8f333b3533972296a4d0557c7b8c48c9cab6ced61",
   "production_predictions_sha256": "ba311f5a6bc73182651bdf5c9744f0ed95921110501ddc93dcfd20a1d06bceff",
   "temperature_sha256": "69d80e5b215e2c1e6872f146fdb7ea5aa95c9fd678ef96208960d50e4e7b635d",
@@ -230,7 +235,7 @@
   "input_length_limit": 16384,
   "original_checkpoint_name": "winner",
   "no_publication_performed": true,
-  "source_bundle_manifest_sha256": "d437bc0149bcb8c9891fbb33c5abc4c2336c3a89ab6cfa0981da5a7f1d19f1f4",
+  "source_bundle_manifest_sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830",
   "normalization_profile_sha256": "be32858d15233e0a3fbee0e4257fb02be0b3439deee4eb9c3f61151df7b73850",
   "public_wrapper_included": true,
   "source_published_bundle_manifest_sha256": "92d7f5be5e1ef21ee682f574b35cc01de4dd8ab16b8aa6edf0dfbdfcfcfabba3",

code/decision_api.py

@@ -55,7 +55,7 @@ def question_row(state, name, question):
         if not isinstance(criteria,dict) or not 2<=len(criteria)<=255:
             raise ValueError('choice requires a mapping of 2..255 criteria')
         if not all(isinstance(k,str) for k in criteria):raise ValueError('Choice keys must be strings')
-        options=[{'key':key,'description':value} for key,value in criteria.items()]
+        options=[{'key':key,'description':key if value is None else value} for key,value in criteria.items()]
     # The question name is used for bookkeeping only; encoders never render id.
     return {'id':name,'state':state,'instructions':question['instructions'],
             'options':options,'task_type':kind,'family':'inference'}

metrics/benchmark.json

@@ -9,7 +9,7 @@
   },
   "metric_design": "User-requested, outcome-informed decision-priority weights; observed regression data, not blind testing.",
   "protocol_sha256": "ab97f652ab890eaefdd5373dd2deeb1e26588815f1f339eee4bcaa610846b0af",
-  "statistics_sha256": "1db772aa92754ce458965f8b449553dde4717d1400dd8fd509ce32d9c76f1d5c",
+  "statistics_sha256": "b319a993ac569ea1fa8da027ccb2195868557e4efe232a5d06091db2884086be",
   "models": {
     "Jev": {
       "panels": {
@@ -10330,11 +10330,11 @@
           ]
         },
         "transfer_v9_test": {
-          "point": 0.6797323135755258,
-          "exact_fraction": "711/1046",
+          "point": 0.6959847036328872,
+          "exact_fraction": "364/523",
           "ci95": [
-            0.6488257294965968,
-            0.7102012834523006
+            0.6660212501997199,
+            0.7249533151094627
           ]
         },
         "original_four_panel_mean": {
@@ -10346,11 +10346,11 @@
           ]
         },
         "weighted_mean": {
-          "point": 0.728414460131567,
-          "exact_fraction": "102402253/140582400",
+          "point": 0.7308523186401712,
+          "exact_fraction": "102744973/140582400",
           "ci95": [
-            0.7132642380063722,
-            0.7430890411518838
+            0.7157461705500261,
+            0.7455873812383725
           ]
         }
       },
@@ -10496,50 +10496,50 @@
         "clean": {
           "requested": 1046,
           "valid_probability_rows": 1046,
-          "correct": 711,
-          "accuracy": 0.6797323135755258,
+          "correct": 728,
+          "accuracy": 0.6959847036328872,
           "full_probability_metrics": {
             "n": 1046,
-            "nll": 0.9375416500577584,
-            "acc": 0.6797323135755258,
-            "ece": 0.10273742120233656,
-            "brier": 0.4350401558940114,
-            "mean_conf": 0.7714702411023553,
-            "confident_error_rate": 0.05449330783938815,
-            "coverage_at_0_9": 0.5239005736137667,
-            "accuracy_at_0_9": 0.8959854014598541,
-            "coverage_at_5pct_error": 0.35181644359464626,
-            "coverage_at_1pct_error": 0.14149139579349904,
-            "aurc": 0.11948583282589058,
-            "error_rate_at_0_9": 0.10401459854014598,
-            "confidence_bias": 0.09173792752682952,
+            "nll": 0.9079076925806586,
+            "acc": 0.6959847036328872,
+            "ece": 0.11803573910263573,
+            "brier": 0.41689550886050714,
+            "mean_conf": 0.814020442735523,
+            "confident_error_rate": 0.058317399617590825,
+            "coverage_at_0_9": 0.5525812619502868,
+            "accuracy_at_0_9": 0.8944636678200693,
+            "coverage_at_5pct_error": 0.36424474187380496,
+            "coverage_at_1pct_error": 0.14340344168260039,
+            "aurc": 0.11088769274123257,
+            "error_rate_at_0_9": 0.10553633217993079,
+            "confidence_bias": 0.11803573910263576,
             "top_bins": {
               "0.9": {
-                "n": 548,
-                "errors": 57,
-                "error_rate": 0.10401459854014598
+                "n": 578,
+                "errors": 61,
+                "error_rate": 0.10553633217993079
               },
               "0.95": {
-                "n": 471,
-                "errors": 38,
-                "error_rate": 0.08067940552016985
+                "n": 491,
+                "errors": 39,
+                "error_rate": 0.07942973523421588
               },
               "0.99": {
-                "n": 303,
-                "errors": 10,
-                "error_rate": 0.033003300330033
+                "n": 314,
+                "errors": 11,
+                "error_rate": 0.03503184713375796
               }
             },
             "selective": {
               "0.5": {
                 "coverage": 0.5,
-                "accuracy": 0.9024856596558317,
-                "confidence_cutoff": 0.9166792767745149
+                "accuracy": 0.9101338432122371,
+                "confidence_cutoff": 0.9301925551074537
               },
               "0.8": {
                 "coverage": 0.8001912045889101,
-                "accuracy": 0.7765830346475507,
-                "confidence_cutoff": 0.5041101725134305
+                "accuracy": 0.7921146953405018,
+                "confidence_cutoff": 0.5957019462220806
               }
             },
             "score_mae": 0.35059676214309027,
@@ -10547,46 +10547,46 @@
           },
           "covered_only_probability_metrics": {
             "n": 1046,
-            "nll": 0.9375416500577584,
-            "acc": 0.6797323135755258,
-            "ece": 0.10273742120233656,
-            "brier": 0.4350401558940114,
-            "mean_conf": 0.7714702411023553,
-            "confident_error_rate": 0.05449330783938815,
-            "coverage_at_0_9": 0.5239005736137667,
-            "accuracy_at_0_9": 0.8959854014598541,
-            "coverage_at_5pct_error": 0.35181644359464626,
-            "coverage_at_1pct_error": 0.14149139579349904,
-            "aurc": 0.11948583282589058,
-            "error_rate_at_0_9": 0.10401459854014598,
-            "confidence_bias": 0.09173792752682952,
+            "nll": 0.9079076925806586,
+            "acc": 0.6959847036328872,
+            "ece": 0.11803573910263573,
+            "brier": 0.41689550886050714,
+            "mean_conf": 0.814020442735523,
+            "confident_error_rate": 0.058317399617590825,
+            "coverage_at_0_9": 0.5525812619502868,
+            "accuracy_at_0_9": 0.8944636678200693,
+            "coverage_at_5pct_error": 0.36424474187380496,
+            "coverage_at_1pct_error": 0.14340344168260039,
+            "aurc": 0.11088769274123257,
+            "error_rate_at_0_9": 0.10553633217993079,
+            "confidence_bias": 0.11803573910263576,
             "top_bins": {
               "0.9": {
-                "n": 548,
-                "errors": 57,
-                "error_rate": 0.10401459854014598
+                "n": 578,
+                "errors": 61,
+                "error_rate": 0.10553633217993079
               },
               "0.95": {
-                "n": 471,
-                "errors": 38,
-                "error_rate": 0.08067940552016985
+                "n": 491,
+                "errors": 39,
+                "error_rate": 0.07942973523421588
               },
               "0.99": {
-                "n": 303,
-                "errors": 10,
-                "error_rate": 0.033003300330033
+                "n": 314,
+                "errors": 11,
+                "error_rate": 0.03503184713375796
               }
             },
             "selective": {
               "0.5": {
                 "coverage": 0.5,
-                "accuracy": 0.9024856596558317,
-                "confidence_cutoff": 0.9166792767745149
+                "accuracy": 0.9101338432122371,
+                "confidence_cutoff": 0.9301925551074537
               },
               "0.8": {
                 "coverage": 0.8001912045889101,
-                "accuracy": 0.7765830346475507,
-                "confidence_cutoff": 0.5041101725134305
+                "accuracy": 0.7921146953405018,
+                "confidence_cutoff": 0.5957019462220806
               }
             },
             "score_mae": 0.35059676214309027,
@@ -10598,33 +10598,33 @@
           "buried_emotion": {
             "requested": 20,
             "valid_probability_rows": 20,
-            "correct": 4,
-            "accuracy": 0.2,
+            "correct": 5,
+            "accuracy": 0.25,
             "full_probability_metrics": {
               "n": 20,
-              "nll": 1.6915727437954797,
-              "acc": 0.2,
-              "ece": 0.09938172959866123,
-              "brier": 0.7953402434802076,
-              "mean_conf": 0.22966574304034268,
+              "nll": 1.4541223366826688,
+              "acc": 0.25,
+              "ece": 0.33647159228575807,
+              "brier": 0.784234954627584,
+              "mean_conf": 0.5836932166148608,
               "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.1,
-              "coverage_at_1pct_error": 0.1,
-              "aurc": 0.611088792957988,
-              "error_rate_at_0_9": null,
-              "confidence_bias": 0.029665743040342668,
+              "coverage_at_0_9": 0.05,
+              "accuracy_at_0_9": 1.0,
+              "coverage_at_5pct_error": 0.05,
+              "coverage_at_1pct_error": 0.05,
+              "aurc": 0.5585412761902699,
+              "error_rate_at_0_9": 0.0,
+              "confidence_bias": 0.33369321661486084,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.95": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.99": {
                   "n": 0,
@@ -10635,41 +10635,41 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.3,
-                  "confidence_cutoff": 0.23240028946100588
+                  "accuracy": 0.5,
+                  "confidence_cutoff": 0.5472205811051023
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.25,
-                  "confidence_cutoff": 0.19576863944530487
+                  "accuracy": 0.3125,
+                  "confidence_cutoff": 0.4314323265237658
                 }
               }
             },
             "covered_only_probability_metrics": {
               "n": 20,
-              "nll": 1.6915727437954797,
-              "acc": 0.2,
-              "ece": 0.09938172959866123,
-              "brier": 0.7953402434802076,
-              "mean_conf": 0.22966574304034268,
+              "nll": 1.4541223366826688,
+              "acc": 0.25,
+              "ece": 0.33647159228575807,
+              "brier": 0.784234954627584,
+              "mean_conf": 0.5836932166148608,
               "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.1,
-              "coverage_at_1pct_error": 0.1,
-              "aurc": 0.611088792957988,
-              "error_rate_at_0_9": null,
-              "confidence_bias": 0.029665743040342668,
+              "coverage_at_0_9": 0.05,
+              "accuracy_at_0_9": 1.0,
+              "coverage_at_5pct_error": 0.05,
+              "coverage_at_1pct_error": 0.05,
+              "aurc": 0.5585412761902699,
+              "error_rate_at_0_9": 0.0,
+              "confidence_bias": 0.33369321661486084,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.95": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.99": {
                   "n": 0,
@@ -10680,13 +10680,13 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.3,
-                  "confidence_cutoff": 0.23240028946100588
+                  "accuracy": 0.5,
+                  "confidence_cutoff": 0.5472205811051023
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.25,
-                  "confidence_cutoff": 0.19576863944530487
+                  "accuracy": 0.3125,
+                  "confidence_cutoff": 0.4314323265237658
                 }
               }
             },
@@ -11475,95 +11475,95 @@
           "emotion": {
             "requested": 80,
             "valid_probability_rows": 80,
-            "correct": 32,
-            "accuracy": 0.4,
+            "correct": 48,
+            "accuracy": 0.6,
             "full_probability_metrics": {
               "n": 80,
-              "nll": 1.652865735056175,
-              "acc": 0.4,
-              "ece": 0.1426948414765427,
-              "brier": 0.7810760834350244,
-              "mean_conf": 0.2911186652438651,
-              "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.0,
-              "coverage_at_1pct_error": 0.0,
-              "aurc": 0.533927612449989,
-              "error_rate_at_0_9": null,
-              "confidence_bias": -0.10888133475613493,
+              "nll": 1.3247643428213003,
+              "acc": 0.6,
+              "ece": 0.15895568320390185,
+              "brier": 0.5466111456851124,
+              "mean_conf": 0.7589556832039019,
+              "confident_error_rate": 0.05,
+              "coverage_at_0_9": 0.3625,
+              "accuracy_at_0_9": 0.8620689655172413,
+              "coverage_at_5pct_error": 0.25,
+              "coverage_at_1pct_error": 0.0375,
+              "aurc": 0.20182699408158983,
+              "error_rate_at_0_9": 0.13793103448275862,
+              "confidence_bias": 0.15895568320390197,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 29,
+                  "errors": 4,
+                  "error_rate": 0.13793103448275862
                 },
                 "0.95": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 19,
+                  "errors": 1,
+                  "error_rate": 0.05263157894736842
                 },
                 "0.99": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 11,
+                  "errors": 1,
+                  "error_rate": 0.09090909090909091
                 }
               },
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.525,
-                  "confidence_cutoff": 0.2595044115111889
+                  "accuracy": 0.85,
+                  "confidence_cutoff": 0.7988864004860878
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.453125,
-                  "confidence_cutoff": 0.21817853532039533
+                  "accuracy": 0.671875,
+                  "confidence_cutoff": 0.5632915364363126
                 }
               }
             },
             "covered_only_probability_metrics": {
               "n": 80,
-              "nll": 1.652865735056175,
-              "acc": 0.4,
-              "ece": 0.1426948414765427,
-              "brier": 0.7810760834350244,
-              "mean_conf": 0.2911186652438651,
-              "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.0,
-              "coverage_at_1pct_error": 0.0,
-              "aurc": 0.533927612449989,
-              "error_rate_at_0_9": null,
-              "confidence_bias": -0.10888133475613493,
+              "nll": 1.3247643428213003,
+              "acc": 0.6,
+              "ece": 0.15895568320390185,
+              "brier": 0.5466111456851124,
+              "mean_conf": 0.7589556832039019,
+              "confident_error_rate": 0.05,
+              "coverage_at_0_9": 0.3625,
+              "accuracy_at_0_9": 0.8620689655172413,
+              "coverage_at_5pct_error": 0.25,
+              "coverage_at_1pct_error": 0.0375,
+              "aurc": 0.20182699408158983,
+              "error_rate_at_0_9": 0.13793103448275862,
+              "confidence_bias": 0.15895568320390197,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 29,
+                  "errors": 4,
+                  "error_rate": 0.13793103448275862
                 },
                 "0.95": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 19,
+                  "errors": 1,
+                  "error_rate": 0.05263157894736842
                 },
                 "0.99": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 11,
+                  "errors": 1,
+                  "error_rate": 0.09090909090909091
                 }
               },
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.525,
-                  "confidence_cutoff": 0.2595044115111889
+                  "accuracy": 0.85,
+                  "confidence_cutoff": 0.7988864004860878
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.453125,
-                  "confidence_cutoff": 0.21817853532039533
+                  "accuracy": 0.671875,
+                  "confidence_cutoff": 0.5632915364363126
                 }
               }
             },
@@ -13243,14 +13243,14 @@
             "coverage": 1.0
           }
         },
-        "clean_task_macro_accuracy": 0.7204166666666667,
+        "clean_task_macro_accuracy": 0.729675925925926,
         "permutation": {
           "requested_pairs": 36,
           "valid_pairs": 36,
-          "both_correct_requested": 0.5833333333333334,
-          "flip_rate": 0.25,
-          "covered_only_flip_rate": 0.25,
-          "half_l1": 0.10492792780864396
+          "both_correct_requested": 0.6388888888888888,
+          "flip_rate": 0.16666666666666666,
+          "covered_only_flip_rate": 0.16666666666666666,
+          "half_l1": 0.07949225370274392
         },
         "unknowable": {
           "requested": 110,
@@ -13378,7 +13378,7 @@
           "truncated_questions": 0
         },
         "complete_unchanged_upstream_report": {
-          "objective": -0.8044143097078044,
+          "objective": -0.7834679467690011,
           "paired_flip": {
             "pairs": 64,
             "flip_rate": 0.828125,
@@ -13399,46 +13399,46 @@
           },
           "clean": {
             "n": 1046,
-            "nll": 0.9375416500577584,
-            "acc": 0.6797323135755258,
-            "ece": 0.10273742120233656,
-            "brier": 0.4350401558940114,
-            "mean_conf": 0.7714702411023553,
-            "confident_error_rate": 0.05449330783938815,
-            "coverage_at_0_9": 0.5239005736137667,
-            "accuracy_at_0_9": 0.8959854014598541,
-            "coverage_at_5pct_error": 0.35181644359464626,
-            "coverage_at_1pct_error": 0.14149139579349904,
-            "aurc": 0.11948583282589058,
-            "error_rate_at_0_9": 0.10401459854014598,
-            "confidence_bias": 0.09173792752682952,
+            "nll": 0.9079076925806586,
+            "acc": 0.6959847036328872,
+            "ece": 0.11803573910263573,
+            "brier": 0.41689550886050714,
+            "mean_conf": 0.814020442735523,
+            "confident_error_rate": 0.058317399617590825,
+            "coverage_at_0_9": 0.5525812619502868,
+            "accuracy_at_0_9": 0.8944636678200693,
+            "coverage_at_5pct_error": 0.36424474187380496,
+            "coverage_at_1pct_error": 0.14340344168260039,
+            "aurc": 0.11088769274123257,
+            "error_rate_at_0_9": 0.10553633217993079,
+            "confidence_bias": 0.11803573910263576,
             "top_bins": {
               "0.9": {
-                "n": 548,
-                "errors": 57,
-                "error_rate": 0.10401459854014598
+                "n": 578,
+                "errors": 61,
+                "error_rate": 0.10553633217993079
               },
               "0.95": {
-                "n": 471,
-                "errors": 38,
-                "error_rate": 0.08067940552016985
+                "n": 491,
+                "errors": 39,
+                "error_rate": 0.07942973523421588
               },
               "0.99": {
-                "n": 303,
-                "errors": 10,
-                "error_rate": 0.033003300330033
+                "n": 314,
+                "errors": 11,
+                "error_rate": 0.03503184713375796
               }
             },
             "selective": {
               "0.5": {
                 "coverage": 0.5,
-                "accuracy": 0.9024856596558317,
-                "confidence_cutoff": 0.9166792767745149
+                "accuracy": 0.9101338432122371,
+                "confidence_cutoff": 0.9301925551074537
               },
               "0.8": {
                 "coverage": 0.8001912045889101,
-                "accuracy": 0.7765830346475507,
-                "confidence_cutoff": 0.5041101725134305
+                "accuracy": 0.7921146953405018,
+                "confidence_cutoff": 0.5957019462220806
               }
             },
             "score_mae": 0.35059676214309027,
@@ -13447,29 +13447,29 @@
           "tasks": {
             "buried_emotion": {
               "n": 20,
-              "nll": 1.6915727437954797,
-              "acc": 0.2,
-              "ece": 0.09938172959866123,
-              "brier": 0.7953402434802076,
-              "mean_conf": 0.22966574304034268,
+              "nll": 1.4541223366826688,
+              "acc": 0.25,
+              "ece": 0.33647159228575807,
+              "brier": 0.784234954627584,
+              "mean_conf": 0.5836932166148608,
               "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.1,
-              "coverage_at_1pct_error": 0.1,
-              "aurc": 0.611088792957988,
-              "error_rate_at_0_9": null,
-              "confidence_bias": 0.029665743040342668,
+              "coverage_at_0_9": 0.05,
+              "accuracy_at_0_9": 1.0,
+              "coverage_at_5pct_error": 0.05,
+              "coverage_at_1pct_error": 0.05,
+              "aurc": 0.5585412761902699,
+              "error_rate_at_0_9": 0.0,
+              "confidence_bias": 0.33369321661486084,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.95": {
-                  "n": 0,
+                  "n": 1,
                   "errors": 0,
-                  "error_rate": null
+                  "error_rate": 0.0
                 },
                 "0.99": {
                   "n": 0,
@@ -13480,13 +13480,13 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.3,
-                  "confidence_cutoff": 0.23240028946100588
+                  "accuracy": 0.5,
+                  "confidence_cutoff": 0.5472205811051023
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.25,
-                  "confidence_cutoff": 0.19576863944530487
+                  "accuracy": 0.3125,
+                  "confidence_cutoff": 0.4314323265237658
                 }
               }
             },
@@ -13854,46 +13854,46 @@
             },
             "emotion": {
               "n": 80,
-              "nll": 1.652865735056175,
-              "acc": 0.4,
-              "ece": 0.1426948414765427,
-              "brier": 0.7810760834350244,
-              "mean_conf": 0.2911186652438651,
-              "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.0,
-              "accuracy_at_0_9": null,
-              "coverage_at_5pct_error": 0.0,
-              "coverage_at_1pct_error": 0.0,
-              "aurc": 0.533927612449989,
-              "error_rate_at_0_9": null,
-              "confidence_bias": -0.10888133475613493,
+              "nll": 1.3247643428213003,
+              "acc": 0.6,
+              "ece": 0.15895568320390185,
+              "brier": 0.5466111456851124,
+              "mean_conf": 0.7589556832039019,
+              "confident_error_rate": 0.05,
+              "coverage_at_0_9": 0.3625,
+              "accuracy_at_0_9": 0.8620689655172413,
+              "coverage_at_5pct_error": 0.25,
+              "coverage_at_1pct_error": 0.0375,
+              "aurc": 0.20182699408158983,
+              "error_rate_at_0_9": 0.13793103448275862,
+              "confidence_bias": 0.15895568320390197,
               "top_bins": {
                 "0.9": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 29,
+                  "errors": 4,
+                  "error_rate": 0.13793103448275862
                 },
                 "0.95": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 19,
+                  "errors": 1,
+                  "error_rate": 0.05263157894736842
                 },
                 "0.99": {
-                  "n": 0,
-                  "errors": 0,
-                  "error_rate": null
+                  "n": 11,
+                  "errors": 1,
+                  "error_rate": 0.09090909090909091
                 }
               },
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.525,
-                  "confidence_cutoff": 0.2595044115111889
+                  "accuracy": 0.85,
+                  "confidence_cutoff": 0.7988864004860878
                 },
                 "0.8": {
                   "coverage": 0.8,
-                  "accuracy": 0.453125,
-                  "confidence_cutoff": 0.21817853532039533
+                  "accuracy": 0.671875,
+                  "confidence_cutoff": 0.5632915364363126
                 }
               }
             },
@@ -15185,46 +15185,46 @@
           "variants": {
             "clean": {
               "n": 1156,
-              "nll": 1.0275747126295693,
-              "acc": 0.6557093425605537,
-              "ece": 0.1271424265612776,
-              "brier": 0.4757092441503964,
-              "mean_conf": 0.7728989400694259,
-              "confident_error_rate": 0.06487889273356401,
-              "coverage_at_0_9": 0.5,
-              "accuracy_at_0_9": 0.870242214532872,
-              "coverage_at_5pct_error": 0.19896193771626297,
-              "coverage_at_1pct_error": 0.12802768166089964,
-              "aurc": 0.14592875492250384,
-              "error_rate_at_0_9": 0.12975778546712802,
-              "confidence_bias": 0.11718959750887226,
-              "top_bins": {
-                "0.9": {
-                  "n": 578,
-                  "errors": 75,
-                  "error_rate": 0.12975778546712802
-                },
-                "0.95": {
-                  "n": 495,
-                  "errors": 51,
-                  "error_rate": 0.10303030303030303
+              "nll": 1.0007605953968302,
+              "acc": 0.6704152249134948,
+              "ece": 0.1409850221700258,
+              "brier": 0.45929116387613567,
+              "mean_conf": 0.8114002470835204,
+              "confident_error_rate": 0.06833910034602077,
+              "coverage_at_0_9": 0.5259515570934256,
+              "accuracy_at_0_9": 0.8700657894736842,
+              "coverage_at_5pct_error": 0.20242214532871972,
+              "coverage_at_1pct_error": 0.12975778546712802,
+              "aurc": 0.1369722577685153,
+              "error_rate_at_0_9": 0.1299342105263158,
+              "confidence_bias": 0.14098502217002562,
+              "top_bins": {
+                "0.9": {
+                  "n": 608,
+                  "errors": 79,
+                  "error_rate": 0.1299342105263158
+                },
+                "0.95": {
+                  "n": 515,
+                  "errors": 52,
+                  "error_rate": 0.10097087378640776
                 },
                 "0.99": {
-                  "n": 317,
-                  "errors": 18,
-                  "error_rate": 0.056782334384858045
+                  "n": 328,
+                  "errors": 19,
+                  "error_rate": 0.057926829268292686
                 }
               },
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.870242214532872,
-                  "confidence_cutoff": 0.9011673130850403
+                  "accuracy": 0.8788927335640139,
+                  "confidence_cutoff": 0.9157804353063602
                 },
                 "0.8": {
-                  "coverage": 0.8001730103806228,
-                  "accuracy": 0.7437837837837837,
-                  "confidence_cutoff": 0.5239011055369641
+                  "coverage": 0.801038062283737,
+                  "accuracy": 0.7591792656587473,
+                  "confidence_cutoff": 0.6007340833161877
                 }
               },
               "score_mae": 0.49813993786469746,
@@ -15232,19 +15232,19 @@
             },
             "none_absent": {
               "n": 36,
-              "nll": 1.9862789221192838,
+              "nll": 2.016479899399725,
               "acc": 0.2777777777777778,
-              "ece": 0.36438230958721407,
-              "brier": 0.971223144388103,
-              "mean_conf": 0.5837732283166512,
+              "ece": 0.43639357499087156,
+              "brier": 1.0193378435955023,
+              "mean_conf": 0.7141713527686493,
               "confident_error_rate": 0.05555555555555555,
               "coverage_at_0_9": 0.1111111111111111,
               "accuracy_at_0_9": 0.5,
               "coverage_at_5pct_error": 0.0,
               "coverage_at_1pct_error": 0.0,
-              "aurc": 0.6812529549976147,
+              "aurc": 0.6622061759903008,
               "error_rate_at_0_9": 0.5,
-              "confidence_bias": 0.30599545053887345,
+              "confidence_bias": 0.4363935749908715,
               "top_bins": {
                 "0.9": {
                   "n": 4,
@@ -15265,39 +15265,39 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.3333333333333333,
-                  "confidence_cutoff": 0.6306419821772263
+                  "accuracy": 0.3888888888888889,
+                  "confidence_cutoff": 0.7284490118427092
                 },
                 "0.8": {
                   "coverage": 0.8055555555555556,
-                  "accuracy": 0.27586206896551724,
-                  "confidence_cutoff": 0.2793320479839753
+                  "accuracy": 0.3448275862068966,
+                  "confidence_cutoff": 0.5772219386282852
                 }
               }
             },
             "none_present": {
               "n": 36,
-              "nll": 1.0208562024633812,
-              "acc": 0.6388888888888888,
-              "ece": 0.1927190351137436,
-              "brier": 0.5059127618759578,
-              "mean_conf": 0.6201323278766935,
+              "nll": 0.9834989953062324,
+              "acc": 0.6666666666666666,
+              "ece": 0.16246789063665687,
+              "brier": 0.42631332906431735,
+              "mean_conf": 0.7475449294183114,
               "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.3333333333333333,
+              "coverage_at_0_9": 0.3888888888888889,
               "accuracy_at_0_9": 1.0,
-              "coverage_at_5pct_error": 0.3888888888888889,
-              "coverage_at_1pct_error": 0.3888888888888889,
-              "aurc": 0.17333125994032075,
+              "coverage_at_5pct_error": 0.5,
+              "coverage_at_1pct_error": 0.5,
+              "aurc": 0.10954859179448669,
               "error_rate_at_0_9": 0.0,
-              "confidence_bias": -0.01875656101219536,
+              "confidence_bias": 0.08087826275164478,
               "top_bins": {
                 "0.9": {
-                  "n": 12,
+                  "n": 14,
                   "errors": 0,
                   "error_rate": 0.0
                 },
                 "0.95": {
-                  "n": 11,
+                  "n": 12,
                   "errors": 0,
                   "error_rate": 0.0
                 },
@@ -15310,39 +15310,39 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.8333333333333334,
-                  "confidence_cutoff": 0.6137714790540333
+                  "accuracy": 1.0,
+                  "confidence_cutoff": 0.8038808995404794
                 },
                 "0.8": {
                   "coverage": 0.8055555555555556,
-                  "accuracy": 0.6551724137931034,
-                  "confidence_cutoff": 0.30252245181248505
+                  "accuracy": 0.7586206896551724,
+                  "confidence_cutoff": 0.514763438695478
                 }
               }
             },
             "permuted": {
               "n": 36,
-              "nll": 0.9278919682020358,
+              "nll": 0.9187906169822195,
               "acc": 0.6666666666666666,
-              "ece": 0.16640228718149716,
-              "brier": 0.4734905986217854,
-              "mean_conf": 0.6290852939746607,
+              "ece": 0.13776171917514016,
+              "brier": 0.4066078577529032,
+              "mean_conf": 0.7543481885836296,
               "confident_error_rate": 0.0,
-              "coverage_at_0_9": 0.3333333333333333,
+              "coverage_at_0_9": 0.4166666666666667,
               "accuracy_at_0_9": 1.0,
-              "coverage_at_5pct_error": 0.3888888888888889,
-              "coverage_at_1pct_error": 0.3888888888888889,
-              "aurc": 0.15567120605578305,
+              "coverage_at_5pct_error": 0.5555555555555556,
+              "coverage_at_1pct_error": 0.5,
+              "aurc": 0.09826261699381092,
               "error_rate_at_0_9": 0.0,
-              "confidence_bias": -0.0375813726920059,
+              "confidence_bias": 0.08768152191696299,
               "top_bins": {
                 "0.9": {
-                  "n": 12,
+                  "n": 15,
                   "errors": 0,
                   "error_rate": 0.0
                 },
                 "0.95": {
-                  "n": 11,
+                  "n": 14,
                   "errors": 0,
                   "error_rate": 0.0
                 },
@@ -15355,13 +15355,13 @@
               "selective": {
                 "0.5": {
                   "coverage": 0.5,
-                  "accuracy": 0.8333333333333334,
-                  "confidence_cutoff": 0.5823386094593029
+                  "accuracy": 1.0,
+                  "confidence_cutoff": 0.8030261229089575
                 },
                 "0.8": {
                   "coverage": 0.8055555555555556,
-                  "accuracy": 0.6896551724137931,
-                  "confidence_cutoff": 0.3432533187230439
+                  "accuracy": 0.7931034482758621,
+                  "confidence_cutoff": 0.5168578524128333
                 }
               }
             }
@@ -15369,52 +15369,52 @@
           "heldout_tasks": {},
           "permutation": {
             "n": 36,
-            "mean_max_delta": 0.08621858211804705,
-            "flip_rate": 0.25
+            "mean_max_delta": 0.07510801427468833,
+            "flip_rate": 0.16666666666666666
           },
           "temperature": 1.0,
           "calibrated_clean": {
             "n": 1046,
-            "nll": 0.9375416500577584,
-            "acc": 0.6797323135755258,
-            "ece": 0.10273742120233656,
-            "brier": 0.4350401558940114,
-            "mean_conf": 0.7714702411023553,
-            "confident_error_rate": 0.05449330783938815,
-            "coverage_at_0_9": 0.5239005736137667,
-            "accuracy_at_0_9": 0.8959854014598541,
-            "coverage_at_5pct_error": 0.35181644359464626,
-            "coverage_at_1pct_error": 0.14149139579349904,
-            "aurc": 0.11948583282589058,
-            "error_rate_at_0_9": 0.10401459854014598,
-            "confidence_bias": 0.09173792752682952,
+            "nll": 0.9079076925806586,
+            "acc": 0.6959847036328872,
+            "ece": 0.11803573910263573,
+            "brier": 0.41689550886050714,
+            "mean_conf": 0.814020442735523,
+            "confident_error_rate": 0.058317399617590825,
+            "coverage_at_0_9": 0.5525812619502868,
+            "accuracy_at_0_9": 0.8944636678200693,
+            "coverage_at_5pct_error": 0.36424474187380496,
+            "coverage_at_1pct_error": 0.14340344168260039,
+            "aurc": 0.11088769274123257,
+            "error_rate_at_0_9": 0.10553633217993079,
+            "confidence_bias": 0.11803573910263576,
             "top_bins": {
               "0.9": {
-                "n": 548,
-                "errors": 57,
-                "error_rate": 0.10401459854014598
+                "n": 578,
+                "errors": 61,
+                "error_rate": 0.10553633217993079
               },
               "0.95": {
-                "n": 471,
-                "errors": 38,
-                "error_rate": 0.08067940552016985
+                "n": 491,
+                "errors": 39,
+                "error_rate": 0.07942973523421588
               },
               "0.99": {
-                "n": 303,
-                "errors": 10,
-                "error_rate": 0.033003300330033
+                "n": 314,
+                "errors": 11,
+                "error_rate": 0.03503184713375796
               }
             },
             "selective": {
               "0.5": {
                 "coverage": 0.5,
-                "accuracy": 0.9024856596558317,
-                "confidence_cutoff": 0.9166792767745149
+                "accuracy": 0.9101338432122371,
+                "confidence_cutoff": 0.9301925551074537
               },
               "0.8": {
                 "coverage": 0.8001912045889101,
-                "accuracy": 0.7765830346475507,
-                "confidence_cutoff": 0.5041101725134305
+                "accuracy": 0.7921146953405018,
+                "confidence_cutoff": 0.5957019462220806
               }
             },
             "score_mae": 0.35059676214309027,
@@ -66859,760 +66859,6 @@
     }
   },
   "paired_comparisons": {
-    "Nox minus Sol": {
-      "old_core": {
-        "delta": 0.09255952380952381,
-        "exact_fraction": "311/3360",
-        "paired_ci95": [
-          0.055577566964285605,
-          0.1305831473214286
-        ]
-      },
-      "v3_core": {
-        "delta": 0.05708333333333333,
-        "exact_fraction": "137/2400",
-        "paired_ci95": [
-          0.0245833333333334,
-          0.08875
-        ]
-      },
-      "v4": {
-        "delta": 0.025,
-        "exact_fraction": "1/40",
-        "paired_ci95": [
-          -0.006288598445143156,
-          0.058030063291139244
-        ]
-      },
-      "v5": {
-        "delta": 0.020833333333333332,
-        "exact_fraction": "1/48",
-        "paired_ci95": [
-          -0.006398928762210043,
-          0.04862658709958234
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.1089866156787763,
-        "exact_fraction": "57/523",
-        "paired_ci95": [
-          0.07972997762299643,
-          0.13779904306220103
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.04886904761904762,
-        "exact_fraction": "821/16800",
-        "paired_ci95": [
-          0.03301257147338517,
-          0.06502318612678312
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.06526168282800691,
-        "exact_fraction": "2293661/35145600",
-        "paired_ci95": [
-          0.04958814518629331,
-          0.08098195894796445
-        ]
-      }
-    },
-    "Nox minus kev-0.8b": {
-      "old_core": {
-        "delta": 0.22864583333333333,
-        "exact_fraction": "439/1920",
-        "paired_ci95": [
-          0.17823567708333335,
-          0.2778273809523808
-        ]
-      },
-      "v3_core": {
-        "delta": 0.095,
-        "exact_fraction": "19/200",
-        "paired_ci95": [
-          0.0633333333333333,
-          0.12791666666666673
-        ]
-      },
-      "v4": {
-        "delta": 0.1125,
-        "exact_fraction": "9/80",
-        "paired_ci95": [
-          0.06741352201257866,
-          0.15727848715504358
-        ]
-      },
-      "v5": {
-        "delta": 0.175,
-        "exact_fraction": "7/40",
-        "paired_ci95": [
-          0.1355393511083524,
-          0.21456521035365878
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.06787762906309751,
-        "exact_fraction": "71/1046",
-        "paired_ci95": [
-          0.03431758137001224,
-          0.10142248780822852
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.15278645833333335,
-        "exact_fraction": "5867/38400",
-        "paired_ci95": [
-          0.13186326764793416,
-          0.1736005686577057
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.1456503943594646,
-        "exact_fraction": "487521/3347200",
-        "paired_ci95": [
-          0.12568410699393018,
-          0.16551635549679958
-        ]
-      }
-    },
-    "Nox minus kev-4b": {
-      "old_core": {
-        "delta": 0.11097470238095238,
-        "exact_fraction": "2983/26880",
-        "paired_ci95": [
-          0.068154761904762,
-          0.1539462425595238
-        ]
-      },
-      "v3_core": {
-        "delta": 0.0325,
-        "exact_fraction": "13/400",
-        "paired_ci95": [
-          0.0033333333333334103,
-          0.06166666666666676
-        ]
-      },
-      "v4": {
-        "delta": -0.028125,
-        "exact_fraction": "-9/320",
-        "paired_ci95": [
-          -0.062443225931676956,
-          0.00613059214504551
-        ]
-      },
-      "v5": {
-        "delta": 0.016666666666666666,
-        "exact_fraction": "1/60",
-        "paired_ci95": [
-          -0.014652826016486136,
-          0.04855467675421749
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.08126195028680688,
-        "exact_fraction": "-85/1046",
-        "paired_ci95": [
-          -0.1091449740378478,
-          -0.05362143534334946
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.03300409226190476,
-        "exact_fraction": "17743/537600",
-        "paired_ci95": [
-          0.0159171480237136,
-          0.05071287650329332
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.027509368171264682,
-        "exact_fraction": "1289111/46860800",
-        "paired_ci95": [
-          0.010595587102736993,
-          0.044668398388024416
-        ]
-      }
-    },
-    "Nox minus kev-9b": {
-      "old_core": {
-        "delta": 0.06253720238095238,
-        "exact_fraction": "1681/26880",
-        "paired_ci95": [
-          0.024514508928571384,
-          0.10111700148809517
-        ]
-      },
-      "v3_core": {
-        "delta": 0.06041666666666667,
-        "exact_fraction": "29/480",
-        "paired_ci95": [
-          0.02416666666666667,
-          0.09708333333333341
-        ]
-      },
-      "v4": {
-        "delta": -0.0765625,
-        "exact_fraction": "-49/640",
-        "paired_ci95": [
-          -0.11406273964723933,
-          -0.040840863146287654
-        ]
-      },
-      "v5": {
-        "delta": 0.027083333333333334,
-        "exact_fraction": "13/480",
-        "paired_ci95": [
-          -0.008081049575002395,
-          0.06178192588510811
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.11281070745697896,
-        "exact_fraction": "-59/523",
-        "paired_ci95": [
-          -0.14365179025963015,
-          -0.08235239701801159
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.018368675595238096,
-        "exact_fraction": "395/21504",
-        "paired_ci95": [
-          0.00015929724577673504,
-          0.03635734921356825
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.009521846262405535,
-        "exact_fraction": "334651/35145600",
-        "paired_ci95": [
-          -0.007595435407328911,
-          0.026801027422374526
-        ]
-      }
-    },
-    "Nox minus Jev": {
-      "old_core": {
-        "delta": 0.0390625,
-        "exact_fraction": "5/128",
-        "paired_ci95": [
-          0.008703497023809521,
-          0.06897321428571423
-        ]
-      },
-      "v3_core": {
-        "delta": -0.14583333333333334,
-        "exact_fraction": "-7/48",
-        "paired_ci95": [
-          -0.18834375000000017,
-          -0.10373958333333338
-        ]
-      },
-      "v4": {
-        "delta": -0.1546875,
-        "exact_fraction": "-99/640",
-        "paired_ci95": [
-          -0.19683761503067482,
-          -0.11406250000000007
-        ]
-      },
-      "v5": {
-        "delta": -0.035416666666666666,
-        "exact_fraction": "-17/480",
-        "paired_ci95": [
-          -0.06684327822543287,
-          -0.0039144015898180265
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.1921606118546845,
-        "exact_fraction": "-201/1046",
-        "paired_ci95": [
-          -0.2229480709592455,
-          -0.16162570888468808
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": -0.07421875,
-        "exact_fraction": "-19/256",
-        "paired_ci95": [
-          -0.09309387400474764,
-          -0.05577344994758131
-        ]
-      },
-      "weighted_mean": {
-        "delta": -0.08207930011153601,
-        "exact_fraction": "-329683/4016640",
-        "paired_ci95": [
-          -0.09913724619038332,
-          -0.06522481396228826
-        ]
-      }
-    },
-    "Nox minus Laya-base": {
-      "old_core": {
-        "delta": 0.26458333333333334,
-        "exact_fraction": "127/480",
-        "paired_ci95": [
-          0.21428478422619038,
-          0.31428664434523806
-        ]
-      },
-      "v3_core": {
-        "delta": 0.16458333333333333,
-        "exact_fraction": "79/480",
-        "paired_ci95": [
-          0.12958333333333344,
-          0.19874999999999993
-        ]
-      },
-      "v4": {
-        "delta": 0.2765625,
-        "exact_fraction": "177/640",
-        "paired_ci95": [
-          0.2216732154792797,
-          0.3312538580246914
-        ]
-      },
-      "v5": {
-        "delta": 0.225,
-        "exact_fraction": "9/40",
-        "paired_ci95": [
-          0.18326279152965536,
-          0.26607286866359453
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.1491395793499044,
-        "exact_fraction": "78/523",
-        "paired_ci95": [
-          0.1103515625,
-          0.18714025122546793
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.23268229166666668,
-        "exact_fraction": "1787/7680",
-        "paired_ci95": [
-          0.20941479053264894,
-          0.2554643867871834
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.218126145235819,
-        "exact_fraction": "4380671/20083200",
-        "paired_ci95": [
-          0.19712453859128337,
-          0.23895587623050776
-        ]
-      }
-    },
-    "Nox minus Laya-multilingual": {
-      "old_core": {
-        "delta": 0.35751488095238093,
-        "exact_fraction": "961/2688",
-        "paired_ci95": [
-          0.3083305431547619,
-          0.40416666666666656
-        ]
-      },
-      "v3_core": {
-        "delta": 0.12875,
-        "exact_fraction": "103/800",
-        "paired_ci95": [
-          0.09583333333333333,
-          0.16166666666666668
-        ]
-      },
-      "v4": {
-        "delta": 0.2828125,
-        "exact_fraction": "181/640",
-        "paired_ci95": [
-          0.2321340794423215,
-          0.33590875589622643
-        ]
-      },
-      "v5": {
-        "delta": 0.28958333333333336,
-        "exact_fraction": "139/480",
-        "paired_ci95": [
-          0.24203304326814726,
-          0.33659227150537624
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.2084130019120459,
-        "exact_fraction": "109/523",
-        "paired_ci95": [
-          0.16923040152963667,
-          0.2462693477059149
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.26466517857142857,
-        "exact_fraction": "11857/44800",
-        "paired_ci95": [
-          0.24161685752065346,
-          0.2875022282879109
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.25656328957252117,
-        "exact_fraction": "12022761/46860800",
-        "paired_ci95": [
-          0.23600300131880433,
-          0.27696035962801896
-        ]
-      }
-    },
-    "Nox minus Qwen3.5-9B": {
-      "old_core": {
-        "delta": 0.09088541666666666,
-        "exact_fraction": "349/3840",
-        "paired_ci95": [
-          0.05066964285714293,
-          0.13192057291666676
-        ]
-      },
-      "v3_core": {
-        "delta": 0.07166666666666667,
-        "exact_fraction": "43/600",
-        "paired_ci95": [
-          0.038333333333333275,
-          0.10499999999999993
-        ]
-      },
-      "v4": {
-        "delta": -0.1078125,
-        "exact_fraction": "-69/640",
-        "paired_ci95": [
-          -0.15406680969159642,
-          -0.0620528650542495
-        ]
-      },
-      "v5": {
-        "delta": 0.06666666666666667,
-        "exact_fraction": "1/15",
-        "paired_ci95": [
-          0.027479948547215406,
-          0.10597745181290595
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.052581261950286805,
-        "exact_fraction": "-55/1046",
-        "paired_ci95": [
-          -0.08596353013081821,
-          -0.01888529584691864
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.0303515625,
-        "exact_fraction": "777/25600",
-        "paired_ci95": [
-          0.01038540473033447,
-          0.050420589249771885
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.031123227374123645,
-        "exact_fraction": "312527/10041600",
-        "paired_ci95": [
-          0.013295818959364762,
-          0.04931011228775257
-        ]
-      }
-    },
-    "Nox minus Lux": {
-      "old_core": {
-        "delta": -0.0020833333333333333,
-        "exact_fraction": "-1/480",
-        "paired_ci95": [
-          -0.03712797619047614,
-          0.03303664434523808
-        ]
-      },
-      "v3_core": {
-        "delta": -0.0008333333333333334,
-        "exact_fraction": "-1/1200",
-        "paired_ci95": [
-          -0.0462499999999999,
-          0.043749999999999956
-        ]
-      },
-      "v4": {
-        "delta": -0.1125,
-        "exact_fraction": "-9/80",
-        "paired_ci95": [
-          -0.15226071147798748,
-          -0.07448441096087467
-        ]
-      },
-      "v5": {
-        "delta": -0.04583333333333333,
-        "exact_fraction": "-11/240",
-        "paired_ci95": [
-          -0.07416056922725012,
-          -0.018749569559228588
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.09464627151051626,
-        "exact_fraction": "-99/1046",
-        "paired_ci95": [
-          -0.12101514862804885,
-          -0.06889873160603051
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": -0.0403125,
-        "exact_fraction": "-129/3200",
-        "paired_ci95": [
-          -0.05905983411444227,
-          -0.02179353585296119
-        ]
-      },
-      "weighted_mean": {
-        "delta": -0.038780274059910774,
-        "exact_fraction": "-48677/1255200",
-        "paired_ci95": [
-          -0.05625030602618532,
-          -0.021731519294624788
-        ]
-      }
-    },
-    "Nox minus Decider": {
-      "old_core": {
-        "delta": 0.18988095238095237,
-        "exact_fraction": "319/1680",
-        "paired_ci95": [
-          0.1443443080357141,
-          0.23571428571428554
-        ]
-      },
-      "v3_core": {
-        "delta": 0.052083333333333336,
-        "exact_fraction": "5/96",
-        "paired_ci95": [
-          0.013333333333333308,
-          0.08999999999999997
-        ]
-      },
-      "v4": {
-        "delta": -0.1296875,
-        "exact_fraction": "-83/640",
-        "paired_ci95": [
-          -0.17179976851851855,
-          -0.08906249999999993
-        ]
-      },
-      "v5": {
-        "delta": 0.01875,
-        "exact_fraction": "3/160",
-        "paired_ci95": [
-          -0.012535333369549428,
-          0.04969926075268807
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.01338432122370937,
-        "exact_fraction": "-7/523",
-        "paired_ci95": [
-          -0.0490208104831727,
-          0.022202716340089176
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.03275669642857143,
-        "exact_fraction": "587/17920",
-        "paired_ci95": [
-          0.013155496386521998,
-          0.05228677715137045
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.05133684586406264,
-        "exact_fraction": "7217057/140582400",
-        "paired_ci95": [
-          0.032083278884638945,
-          0.07037086432355429
-        ]
-      }
-    },
-    "Nox minus Qwen3.5-2B": {
-      "old_core": {
-        "delta": 0.25877976190476193,
-        "exact_fraction": "1739/6720",
-        "paired_ci95": [
-          0.21566220238095246,
-          0.30174851190476193
-        ]
-      },
-      "v3_core": {
-        "delta": 0.12791666666666668,
-        "exact_fraction": "307/2400",
-        "paired_ci95": [
-          0.0979166666666666,
-          0.15999999999999992
-        ]
-      },
-      "v4": {
-        "delta": 0.053125,
-        "exact_fraction": "17/320",
-        "paired_ci95": [
-          0.0011595366002794232,
-          0.10381534584912663
-        ]
-      },
-      "v5": {
-        "delta": 0.13958333333333334,
-        "exact_fraction": "67/480",
-        "paired_ci95": [
-          0.09772693403789712,
-          0.18099321705426352
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.11663479923518165,
-        "exact_fraction": "61/523",
-        "paired_ci95": [
-          0.08030488916814263,
-          0.15300807906502473
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.14485119047619047,
-        "exact_fraction": "4867/33600",
-        "paired_ci95": [
-          0.12338022237889658,
-          0.1658467989420549
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.15601456512337247,
-        "exact_fraction": "10966451/70291200",
-        "paired_ci95": [
-          0.13723175354986636,
-          0.17471297097636712
-        ]
-      }
-    },
-    "Nox minus Qwen3.5-4B": {
-      "old_core": {
-        "delta": 0.13113839285714285,
-        "exact_fraction": "235/1792",
-        "paired_ci95": [
-          0.09270647321428567,
-          0.16990420386904764
-        ]
-      },
-      "v3_core": {
-        "delta": 0.08458333333333333,
-        "exact_fraction": "203/2400",
-        "paired_ci95": [
-          0.05416666666666664,
-          0.11541666666666661
-        ]
-      },
-      "v4": {
-        "delta": -0.0890625,
-        "exact_fraction": "-57/640",
-        "paired_ci95": [
-          -0.1342959349117049,
-          -0.045753734276729595
-        ]
-      },
-      "v5": {
-        "delta": 0.06458333333333334,
-        "exact_fraction": "31/480",
-        "paired_ci95": [
-          0.028211805555555566,
-          0.10191039756413821
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.008604206500956023,
-        "exact_fraction": "-9/1046",
-        "paired_ci95": [
-          -0.040737677035593334,
-          0.022878932316491855
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.04781063988095238,
-        "exact_fraction": "25703/537600",
-        "paired_ci95": [
-          0.028947119327501308,
-          0.06683696978735856
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.05552484521533279,
-        "exact_fraction": "975727/17572800",
-        "paired_ci95": [
-          0.03826384999615495,
-          0.0725083343854517
-        ]
-      }
-    },
-    "Sol minus Nox": {
-      "old_core": {
-        "delta": -0.09255952380952381,
-        "exact_fraction": "-311/3360",
-        "paired_ci95": [
-          -0.13058314732142862,
-          -0.055577566964285646
-        ]
-      },
-      "v3_core": {
-        "delta": -0.05708333333333333,
-        "exact_fraction": "-137/2400",
-        "paired_ci95": [
-          -0.08875,
-          -0.0245833333333334
-        ]
-      },
-      "v4": {
-        "delta": -0.025,
-        "exact_fraction": "-1/40",
-        "paired_ci95": [
-          -0.058030063291139244,
-          0.006288598445143156
-        ]
-      },
-      "v5": {
-        "delta": -0.020833333333333332,
-        "exact_fraction": "-1/48",
-        "paired_ci95": [
-          -0.048626587099582376,
-          0.0063989287622100415
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": -0.1089866156787763,
-        "exact_fraction": "-57/523",
-        "paired_ci95": [
-          -0.13779904306220103,
-          -0.07972997762299644
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": -0.04886904761904762,
-        "exact_fraction": "-821/16800",
-        "paired_ci95": [
-          -0.06502318612678314,
-          -0.033012571473385176
-        ]
-      },
-      "weighted_mean": {
-        "delta": -0.06526168282800691,
-        "exact_fraction": "-2293661/35145600",
-        "paired_ci95": [
-          -0.08098195894796448,
-          -0.049588145186293314
-        ]
-      }
-    },
     "Sol minus kev-0.8b": {
       "old_core": {
         "delta": 0.1360863095238095,
@@ -68251,64 +67497,6 @@
         ]
       }
     },
-    "Lux minus Nox": {
-      "old_core": {
-        "delta": 0.0020833333333333333,
-        "exact_fraction": "1/480",
-        "paired_ci95": [
-          -0.033036644345238085,
-          0.03712797619047614
-        ]
-      },
-      "v3_core": {
-        "delta": 0.0008333333333333334,
-        "exact_fraction": "1/1200",
-        "paired_ci95": [
-          -0.043749999999999956,
-          0.0462499999999999
-        ]
-      },
-      "v4": {
-        "delta": 0.1125,
-        "exact_fraction": "9/80",
-        "paired_ci95": [
-          0.07448441096087466,
-          0.15226071147798748
-        ]
-      },
-      "v5": {
-        "delta": 0.04583333333333333,
-        "exact_fraction": "11/240",
-        "paired_ci95": [
-          0.018749569559228584,
-          0.07416056922725009
-        ]
-      },
-      "transfer_v9_test": {
-        "delta": 0.09464627151051626,
-        "exact_fraction": "99/1046",
-        "paired_ci95": [
-          0.0688987316060305,
-          0.12101514862804882
-        ]
-      },
-      "original_four_panel_mean": {
-        "delta": 0.0403125,
-        "exact_fraction": "129/3200",
-        "paired_ci95": [
-          0.02179353585296119,
-          0.05905983411444227
-        ]
-      },
-      "weighted_mean": {
-        "delta": 0.038780274059910774,
-        "exact_fraction": "48677/1255200",
-        "paired_ci95": [
-          0.021731519294624784,
-          0.05625030602618531
-        ]
-      }
-    },
     "Lux minus Sol": {
       "old_core": {
         "delta": 0.09464285714285714,
@@ -68946,6 +68134,180 @@
           0.11044819701768911
         ]
       }
+    },
+    "Nox minus kev-4b": {
+      "old_core": {
+        "delta": 0.11097470238095238,
+        "exact_fraction": "2983/26880",
+        "paired_ci95": [
+          0.068154761904762,
+          0.1539462425595238
+        ]
+      },
+      "v3_core": {
+        "delta": 0.0325,
+        "exact_fraction": "13/400",
+        "paired_ci95": [
+          0.0033333333333334103,
+          0.06166666666666676
+        ]
+      },
+      "v4": {
+        "delta": -0.028125,
+        "exact_fraction": "-9/320",
+        "paired_ci95": [
+          -0.062443225931676956,
+          0.00613059214504551
+        ]
+      },
+      "v5": {
+        "delta": 0.016666666666666666,
+        "exact_fraction": "1/60",
+        "paired_ci95": [
+          -0.014652826016486136,
+          0.04855467675421749
+        ]
+      },
+      "transfer_v9_test": {
+        "delta": -0.06500956022944551,
+        "exact_fraction": "-34/523",
+        "paired_ci95": [
+          -0.09099616858237547,
+          -0.03925188928140224
+        ]
+      },
+      "original_four_panel_mean": {
+        "delta": 0.03300409226190476,
+        "exact_fraction": "17743/537600",
+        "paired_ci95": [
+          0.0159171480237136,
+          0.05071287650329332
+        ]
+      },
+      "weighted_mean": {
+        "delta": 0.029947226679868887,
+        "exact_fraction": "1403351/46860800",
+        "paired_ci95": [
+          0.013201872388473554,
+          0.047038737697081195
+        ]
+      }
+    },
+    "Nox minus Qwen3.5-4B": {
+      "old_core": {
+        "delta": 0.13113839285714285,
+        "exact_fraction": "235/1792",
+        "paired_ci95": [
+          0.09270647321428567,
+          0.16990420386904764
+        ]
+      },
+      "v3_core": {
+        "delta": 0.08458333333333333,
+        "exact_fraction": "203/2400",
+        "paired_ci95": [
+          0.05416666666666664,
+          0.11541666666666661
+        ]
+      },
+      "v4": {
+        "delta": -0.0890625,
+        "exact_fraction": "-57/640",
+        "paired_ci95": [
+          -0.1342959349117049,
+          -0.045753734276729595
+        ]
+      },
+      "v5": {
+        "delta": 0.06458333333333334,
+        "exact_fraction": "31/480",
+        "paired_ci95": [
+          0.028211805555555566,
+          0.10191039756413821
+        ]
+      },
+      "transfer_v9_test": {
+        "delta": 0.0076481835564053535,
+        "exact_fraction": "4/523",
+        "paired_ci95": [
+          -0.022396190701526143,
+          0.03838776086034498
+        ]
+      },
+      "original_four_panel_mean": {
+        "delta": 0.04781063988095238,
+        "exact_fraction": "25703/537600",
+        "paired_ci95": [
+          0.028947119327501308,
+          0.06683696978735856
+        ]
+      },
+      "weighted_mean": {
+        "delta": 0.05796270372393699,
+        "exact_fraction": "1018567/17572800",
+        "paired_ci95": [
+          0.04083969253575949,
+          0.07488554488575702
+        ]
+      }
+    },
+    "Nox minus Decider": {
+      "old_core": {
+        "delta": 0.18988095238095237,
+        "exact_fraction": "319/1680",
+        "paired_ci95": [
+          0.1443443080357141,
+          0.23571428571428554
+        ]
+      },
+      "v3_core": {
+        "delta": 0.052083333333333336,
+        "exact_fraction": "5/96",
+        "paired_ci95": [
+          0.013333333333333308,
+          0.08999999999999997
+        ]
+      },
+      "v4": {
+        "delta": -0.1296875,
+        "exact_fraction": "-83/640",
+        "paired_ci95": [
+          -0.17179976851851855,
+          -0.08906249999999993
+        ]
+      },
+      "v5": {
+        "delta": 0.01875,
+        "exact_fraction": "3/160",
+        "paired_ci95": [
+          -0.012535333369549428,
+          0.04969926075268807
+        ]
+      },
+      "transfer_v9_test": {
+        "delta": 0.0028680688336520078,
+        "exact_fraction": "3/1046",
+        "paired_ci95": [
+          -0.030160845668367766,
+          0.03653934071222337
+        ]
+      },
+      "original_four_panel_mean": {
+        "delta": 0.03275669642857143,
+        "exact_fraction": "587/17920",
+        "paired_ci95": [
+          0.013155496386521998,
+          0.05228677715137045
+        ]
+      },
+      "weighted_mean": {
+        "delta": 0.05377470437266685,
+        "exact_fraction": "7559777/140582400",
+        "paired_ci95": [
+          0.03467000712516513,
+          0.07268914300878532
+        ]
+      }
     }
   },
   "method": {
@@ -68973,10 +68335,10 @@
     "transfer_accuracy": "Upstream ordered probability argmax, clean knowable records, micro average. Refused/missing probabilities count wrong; unknown evidence is not accuracy.",
     "transfer_clusters": "Union IDs/group/parent/control/pair plus exact state SHA and ordered-request IDs, built over all variants then scored only clean knowable rows. Complete components jointly resampled; variable sampled denominator. No whole-family grouping.",
     "pairing": "Identical sampled source components across all models; five panels sampled independently. Original four marginal CIs preserved.",
-    "accuracy_arithmetic": "Exact rational integer correct/requested. Product weights were explicitly chosen after observing earlier outcomes; no claim of outcome-independent weighting.",
+    "accuracy_arithmetic": "Exact rational integer correct/requested; weights not fitted to outcomes.",
     "interval_role": "Descriptive paired95% intervals; no positive-CI release requirement invented.",
     "diagnostics": "Full per-task/proper-score/order/unknown-evidence results remain separate; no accuracy/latency/ECE unit mixing.",
-    "exposure": "Observed regression data; latest30/25/15/15/15 weighting is outcome-informed user steering. Identical predictions, no test-based checkpoint reselection or calibration."
+    "exposure": "Observed regression tests; weighting user-approved after earlier results; not a blind prospective benchmark."
   },
   "prior_weighted_scores": {
     "Jev": {
@@ -68996,12 +68358,9 @@
       ]
     },
     "Nox": {
-      "point": 0.7208999597389147,
-      "exact_fraction": "202691693/281164800",
-      "ci95": [
-        0.706127064217637,
-        0.7352055352744082
-      ]
+      "point": 0.724150437750387,
+      "exact_fraction": "203605613/281164800",
+      "interval_status": "Not recomputed for alternate weighting; only the point is displayed."
     },
     "kev-9b": {
       "point": 0.7201455089684057,
@@ -69086,6 +68445,6 @@
   },
   "hosted_frontier": "All official calls returned jev-1.13.0; closed weight identity/input retention cannot be independently inspected.",
   "size_note": "Qwen3.5 family size labels name their official parent checkpoint. Lux deployed text+decision weights total7,940,895,744 parameters; no vision or generativeLMhead in its serving bundle.",
-  "presentation_amendment_sha256": "f078ecb12e7f4f2baaab522d07af6695c18d9c4669a5f61623e85c93b910fd1f",
-  "presentation_note": "Public display roster changed only; frozen protocol, complete internal evidence and every retained numeric result are unchanged."
+  "presentation_note": "Latest released result per model. Nox uses its qualified Choice null-description rendering; other displayed model results are unchanged. No historical model rows.",
+  "update_kind": "Nox SystemOne Choice semantics: null descriptions use their key text; model weights, tokenizer and temperature unchanged."
 }

metrics/evaluation-provenance.json

@@ -1,145 +1,148 @@
 {
-  "statistics_sha256": "1db772aa92754ce458965f8b449553dde4717d1400dd8fd509ce32d9c76f1d5c",
-  "manifest_sha256": "3199a40b2719cad9b1e815629cf5ebad923a9a399188c1199a9d27e918bda9e3",
+  "frozen_protocol_sha256": "ab97f652ab890eaefdd5373dd2deeb1e26588815f1f339eee4bcaa610846b0af",
+  "statistics_sha256": "b319a993ac569ea1fa8da027ccb2195868557e4efe232a5d06091db2884086be",
+  "manifest_sha256": "cf455486f8eba2b7158bbb4b5bcdcd8a4bd4daee4f05184b1df61b5f70ef2b04",
   "qualified_runtime": {
     "python": "3.12.13",
     "numpy": "2.3.5"
   },
-  "source_sha256": "b00c60f21ace9bd258a1a79673b8cdbce3101ada2fed5547811f70bf030fcd91",
+  "source_sha256": "f2f72fe161944599b1114df29ff095786bf661cca59afd9d784b7ca6d1430c99",
   "model_manifest": {
-    "Nox": {
+    "Jev": {
       "panels": {
         "old_core": {
-          "path": "results/expanded/nox-retention-v1/candidate/job-0/worker/old_core/normalized.jsonl",
-          "sha256": "9b6b7db907e8e458cd33982c0f7c45e5052c3248285d927990f160d5570c8976"
+          "path": "eval/heldout/official-predictions.jsonl",
+          "sha256": "069e48e1fdf1b5f3d95cc0097e4179ac3faab97bb357b0ce897cfb463b87cbaf",
+          "bytes": 185263
         },
         "v3_core": {
-          "path": "results/expanded/nox-retention-v1/candidate/job-0/worker/v3_core/normalized.jsonl",
-          "sha256": "49575aef2261c191e289d781857664c57fd6181701c03f02515292f2fa31b40c"
+          "path": "eval/heldout/v3/official/normalized/core.jsonl",
+          "sha256": "54707354bfe5d96b5de56e6dc1b38cba0a610eea516b28fbff013b94f422b397",
+          "bytes": 368716
         },
         "v4": {
-          "path": "results/expanded/nox-retention-v1/candidate/job-0/worker/v4/normalized.jsonl",
-          "sha256": "9d85618b1a05c2ed68f6a188dadf19a64d6ede6d5a1b0ef58afa27a13906238b"
+          "path": "eval/heldout/v4/official/normalized/core.jsonl",
+          "sha256": "d6583f903e974d70747fae2944014828954dc4d7344ceddcbb2cd1b06fbfc799",
+          "bytes": 183671
         },
         "v5": {
-          "path": "results/expanded/nox-retention-v1/candidate/job-0/worker/v5/normalized.jsonl",
-          "sha256": "5eb7faaa45a22dad414af1363d6087b8c379c86599df3c14483a3a648ca6bfd5"
+          "path": "eval/heldout/v5/official/normalized/core.jsonl",
+          "sha256": "f883164c4c83db8e42ebe03f7e165afbd2f81f2b541bcf8008a2cc06a60e1ee7",
+          "bytes": 200316
         }
       },
       "transfer_rows": {
-        "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Nox/v9-transfer-v9-test-published-temperature-rows.json",
-        "sha256": "77b92a18433b260ac0e6f4e0b6c9ec3a94e16aaed870a4078b1d7f32ef18b169"
+        "path": "analysis/decision-benchmark-v3-official/scored/ROWS.json",
+        "sha256": "be29a7f3804ad84993ffb06f05ada7e2d8ed25317504c4fc17fe5dc306fd788c"
       },
       "evidence": [
         {
-          "path": "results/expanded/nox-retention-v1/QUALITY-COLLECTION.json",
-          "sha256": "7608258282ec1e9731e4d2ea4dee7206fda8900055dc64f9193d2566172ca86a"
+          "path": "eval/expanded-public-extension-v1/COMPARATOR-REGISTRY.json",
+          "sha256": "b80be4270721cbeeac9e28beb10c6f19f75c1cbf2553e38d3d59e3a42d56a91c"
+        },
+        {
+          "path": "analysis/decision-benchmark-v3-official/scored/REPORT.json",
+          "sha256": "2fdbaf7db1e4b877762e33a87b23a236a065eaa3078b14c2d7d579ad44f9ea24"
         },
         {
-          "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Nox/REPORT.json",
-          "sha256": "8e78910746c9e78862b3aadd94490054a681f3ad013b43ec19bfbc2434111ed7"
+          "path": "analysis/decision-benchmark-v3-official/INDEPENDENT-PROJECTION-REVIEW.json",
+          "sha256": "b9506539e2c4a863aef16f7dab0517c175e85863416bc622c60df05489a9eb52"
         }
       ]
     },
-    "Sol": {
+    "Lux": {
       "panels": {
         "old_core": {
-          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/old_core/normalized.jsonl",
-          "sha256": "5fc0b3e49a5de576f7621839130855c178b12d942dc357aaf6b4dd9735e0a6a8"
+          "path": "results/lux-new-host-v1/quality/lux/old_core/normalized.jsonl",
+          "sha256": "a659a7c7fef2871bbf806e842ea8ca81f645dcf610f0f2f6622d8d32e76a0836"
         },
         "v3_core": {
-          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v3_core/normalized.jsonl",
-          "sha256": "b2c99fbd739b849a66a55929a658de4517a7ffaa4496255566e6972ba28d4f15"
+          "path": "results/lux-new-host-v1/quality/lux/v3_core/normalized.jsonl",
+          "sha256": "da8903a07e36a9687f6ee1d0fd8dbc684e61c49ef5fea30c346e46bff5a90e94"
         },
         "v4": {
-          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v4/normalized.jsonl",
-          "sha256": "3c28302750e02d4cead0aacd66a323734e773f726c4c41db658dd6461251b28b"
+          "path": "results/lux-new-host-v1/quality/lux/v4/normalized.jsonl",
+          "sha256": "26cd7482693ec1ffa1a93a156abfc88378e4872e320f3b9b775763635086d488"
         },
         "v5": {
-          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v5/normalized.jsonl",
-          "sha256": "6ddb0496012b596ea122afa4db4c4158e57dfb934e9ae29eb006756b652e31e1"
+          "path": "results/lux-new-host-v1/quality/lux/v5/normalized.jsonl",
+          "sha256": "759c773e629a16317271de850780a8734533c795046122f850914c4482d1e72b"
         }
       },
       "transfer_rows": {
-        "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Sol-attempt02/v9-transfer-v9-test-published-temperature-rows.json",
-        "sha256": "b05799417cb3715de633135c58f713960b75bdb9ed30eb8912a197d33ee3d85b"
+        "path": "analysis/decision-benchmark-v3-baselines/Lux/ROWS.json",
+        "sha256": "bafb55fc86762061e537382360eda14a8a8f8c5e4e111c2fda1f82cd6c06793f"
       },
       "evidence": [
         {
-          "path": "results/expanded/sol-composition-v2/QUALITY-COLLECTION.json",
-          "sha256": "9b77aac87b85cbf3a74e86af4a2b9a03fd9e1ed854356a92ff54ccd0d47a16f4"
+          "path": "results/lux-new-host-v1/quality/lux/COMPLETE.json",
+          "sha256": "9600ff792cbda13af5505c804cf63af94d2ee895f2912f0645a7d493ae1b7a9c"
         },
         {
-          "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Sol-attempt02/REPORT.json",
-          "sha256": "5d8ebb339293687cec9089f54fb96aabf3330f16111487a49489f5e52bbb6f29"
+          "path": "results/lux-new-host-v1/quality/lux/metadata.json",
+          "sha256": "0d4509fae7eb3630118eaebc24e75a7b02f717d9a0ad74811fa27358c33c9cc5"
+        },
+        {
+          "path": "analysis/decision-benchmark-v3-baselines/Lux-projected/COMPLETE.json",
+          "sha256": "f58d02df78f297d4feebe39facfac072951111d8d6a19e2c1069f1b90250c622"
+        },
+        {
+          "path": "analysis/decision-benchmark-v3-baselines/Lux/REPORT.json",
+          "sha256": "f31703c342efe043a1e1534da189808bb7e0de0641fbaf3f91faa4a1f928a299"
+        },
+        {
+          "path": "analysis/decoder4b/lux9b-training-v1/CHECKPOINT-HELDOUT-COLLECTED.json",
+          "sha256": "d1c42515ab225b27ea011a6cf2262fe12dd7001e3b4d2cb64a23e7642c7a0d91"
         }
       ]
     },
-    "kev-0.8b": {
+    "Nox": {
       "panels": {
         "old_core": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-old_core-normalized.jsonl",
-          "sha256": "b492b4512ac164f83a319f31bbc9e086cae12f87ffe2c14f307481ffb240807c"
+          "path": "results/nox-null-description-v1/full-regression/old_core/normalized.jsonl",
+          "sha256": "dbd18273dbbb9d09ba37d19dfedb0fcd6b64841d528632f04d39fd7955799967"
         },
         "v3_core": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v3_core-normalized.jsonl",
-          "sha256": "70cbfc69dfc72ac0c5ae01305ec63a9f31129c115b7068c6824cd22ead013d68"
+          "path": "results/nox-null-description-v1/full-regression/v3_core/normalized.jsonl",
+          "sha256": "5286d1c3a78649e3284a018b1518f94cf32d8f0ad43a49c6a14c993524a3aff8"
         },
         "v4": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v4-normalized.jsonl",
-          "sha256": "761dfd69acbface101f2c21fd27d2db684c5c8b8cc425890a075729ae1f7db24"
+          "path": "results/nox-null-description-v1/full-regression/v4/normalized.jsonl",
+          "sha256": "077b752f4866bd0fc83db9b31c6227fd5015ce3585d2138c1a7a803c433d472b"
         },
         "v5": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v5-normalized.jsonl",
-          "sha256": "51b6fc5f826ca8db44df43885f34d857ddd5b5a66453bbdc1a83fcf488928bdb"
+          "path": "results/nox-null-description-v1/full-regression/v5/normalized.jsonl",
+          "sha256": "d57f07df59ff9b0e6ace0658e7917746e2b8d3bf518309a3858045cd175d79df"
         }
       },
       "transfer_rows": {
-        "path": "analysis/kev-reciprocal-v1/reports/kev-0.8b/v9-transfer-v9-test-published-temperature-rows.json",
-        "sha256": "d6abed88001aeac630219466f02c1aad89fcbd354b3ae533507f3c2f7fe42898"
+        "path": "analysis/nox-null-description-candidate-v4/transfer/ROWS.json",
+        "sha256": "7cee9e036ee53a3c41918cfafc73ad0d68f16b266abcae43077d1d69247afdfe"
       },
       "evidence": [
         {
-          "path": "analysis/kev-reciprocal-v1/reports/kev-0.8b/REPORT.json",
-          "sha256": "6760dcbaa7d85e6c2e893e4c80ab8e66e0b84d7cfa26ae9e25909f48a05f8de4"
+          "path": "results/nox-null-description-v1/COMPLETE.json",
+          "sha256": "b813aa5db8c43b728b94abbaef20cefc8d3d84b51f7873ca46b3afb59aab6fe4"
         },
         {
-          "path": "analysis/kev-reciprocal-v1/our-panels/POINTS.json",
-          "sha256": "a3aa6d501a14df1cf5e7c98ecc6161c88483801310c56913465059d28c4d88ea"
-        }
-      ]
-    },
-    "kev-4b": {
-      "panels": {
-        "old_core": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-old_core-normalized.jsonl",
-          "sha256": "41dfbb7c9c4381effc219f9123b51bd4d688b3e746cec974ef930730b9dfd6b1"
+          "path": "results/nox-null-description-v1/full-regression/COMPLETE.json",
+          "sha256": "ad80e5bfb2558ee3b8fb1fc80617fcefbc1fbb8e58b8155e3cf5dfde38528754"
         },
-        "v3_core": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v3_core-normalized.jsonl",
-          "sha256": "2ec54da92c4e72412e89b3c07013b54729f5899ba8542eeadf962b04a1f3a4b8"
+        {
+          "path": "results/nox-null-description-v1/transfer-v9/COMPLETE.json",
+          "sha256": "93a8767f7245b93764e527e616f236b5fa46b24a087a7e8ff66d35eee9f9fd0f"
         },
-        "v4": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v4-normalized.jsonl",
-          "sha256": "88c576c3d62ec11686dd5ebf360971a9087779e45a346897e2badf874a754a12"
+        {
+          "path": "results/nox-null-description-v1/full-regression-ACTUAL-EXIT.json",
+          "sha256": "dca6e05615c9c82a2d506438fb4e7c9abfe03053661515ffc9a4c5b176fd3ed8"
         },
-        "v5": {
-          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v5-normalized.jsonl",
-          "sha256": "d33bbb60134314dcb550060176ed0a3ae91ec0c6b033cceb40e3255b33666898"
-        }
-      },
-      "transfer_rows": {
-        "path": "analysis/kev-reciprocal-v1/reports/kev-4b/v9-transfer-v9-test-published-temperature-rows.json",
-        "sha256": "c2cb16a936be4c10ca6b1870c69426ddd6b51629c401f71ae1da652b3bd2ac7e"
-      },
-      "evidence": [
         {
-          "path": "analysis/kev-reciprocal-v1/reports/kev-4b/REPORT.json",
-          "sha256": "366fb0d27a4d732e5e270c264521d9a93ccabc4f0ddf5b7d3e9f1ecc5f460406"
+          "path": "results/nox-null-description-v1/transfer-v9-ACTUAL-EXIT.json",
+          "sha256": "09c4c05549253c386f72b52e8ad4fbbc47d05af59715caebf9d01a6e8734d0af"
         },
         {
-          "path": "analysis/kev-reciprocal-v1/our-panels/POINTS.json",
-          "sha256": "a3aa6d501a14df1cf5e7c98ecc6161c88483801310c56913465059d28c4d88ea"
+          "path": "analysis/nox-null-description-candidate-v4/transfer/REPORT.json",
+          "sha256": "f02e447787c0c5afcb8acbd06408407719ebfc191e8501e8a887cd96c150f81b"
         }
       ]
     },
@@ -177,129 +180,37 @@
         }
       ]
     },
-    "Jev": {
-      "panels": {
-        "old_core": {
-          "path": "eval/heldout/official-predictions.jsonl",
-          "sha256": "069e48e1fdf1b5f3d95cc0097e4179ac3faab97bb357b0ce897cfb463b87cbaf",
-          "bytes": 185263
-        },
-        "v3_core": {
-          "path": "eval/heldout/v3/official/normalized/core.jsonl",
-          "sha256": "54707354bfe5d96b5de56e6dc1b38cba0a610eea516b28fbff013b94f422b397",
-          "bytes": 368716
-        },
-        "v4": {
-          "path": "eval/heldout/v4/official/normalized/core.jsonl",
-          "sha256": "d6583f903e974d70747fae2944014828954dc4d7344ceddcbb2cd1b06fbfc799",
-          "bytes": 183671
-        },
-        "v5": {
-          "path": "eval/heldout/v5/official/normalized/core.jsonl",
-          "sha256": "f883164c4c83db8e42ebe03f7e165afbd2f81f2b541bcf8008a2cc06a60e1ee7",
-          "bytes": 200316
-        }
-      },
-      "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-official/scored/ROWS.json",
-        "sha256": "be29a7f3804ad84993ffb06f05ada7e2d8ed25317504c4fc17fe5dc306fd788c"
-      },
-      "evidence": [
-        {
-          "path": "eval/expanded-public-extension-v1/COMPARATOR-REGISTRY.json",
-          "sha256": "b80be4270721cbeeac9e28beb10c6f19f75c1cbf2553e38d3d59e3a42d56a91c"
-        },
-        {
-          "path": "analysis/decision-benchmark-v3-official/scored/REPORT.json",
-          "sha256": "2fdbaf7db1e4b877762e33a87b23a236a065eaa3078b14c2d7d579ad44f9ea24"
-        },
-        {
-          "path": "analysis/decision-benchmark-v3-official/INDEPENDENT-PROJECTION-REVIEW.json",
-          "sha256": "b9506539e2c4a863aef16f7dab0517c175e85863416bc622c60df05489a9eb52"
-        }
-      ]
-    },
-    "Laya-base": {
+    "kev-4b": {
       "panels": {
-        "v4": {
-          "path": "results/public-roster-baselines-v1/laya-english/worker/v4/normalized.jsonl",
-          "sha256": "31da1d4140da69d2677172707eb301843cf8fdfb43200cba93be910321f922ba"
-        },
-        "v5": {
-          "path": "results/public-roster-baselines-v1/laya-english/worker/v5/normalized.jsonl",
-          "sha256": "159614ceede814b578e1e025836a595c8c71a9bec4896caf6454d7bf54c4c44d"
-        },
         "old_core": {
-          "path": "eval/heldout/primary-v2/laya-english/core/normalized.jsonl",
-          "sha256": "850643d2daeb47a6909985003736cab0f1f0e5a87631a35e27192f0733efd756"
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-old_core-normalized.jsonl",
+          "sha256": "41dfbb7c9c4381effc219f9123b51bd4d688b3e746cec974ef930730b9dfd6b1"
         },
         "v3_core": {
-          "path": "eval/heldout/v3/baselines/laya-english/core/normalized.jsonl",
-          "sha256": "dec19caeb8074e25289e43a367392172da21f0c2ccdb403a059e2e4a2364a64b"
-        }
-      },
-      "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-baselines/laya-english/ROWS.json",
-        "sha256": "19138909259400facb5d3d5402d0d736f870695fe3ee46019f2e438088c57275"
-      },
-      "evidence": [
-        {
-          "path": "results/public-roster-baselines-v1/laya-english/worker/COMPLETE.json",
-          "sha256": "cef0da797ae14cebaf1e4b9edf9b16c3cd9a412a30435d7d2517b8ba6dbe8b98"
-        },
-        {
-          "path": "analysis/public-roster-discovery-v1/ACTUAL-EVALUATION-COMPLETE.json",
-          "sha256": "686834a8289d3771908530a95c2ff85df71e61dbb1ce7ea95b7c9550a23ba82f"
-        },
-        {
-          "path": "analysis/decision-benchmark-v3-baselines/laya-english/REPORT.json",
-          "sha256": "cb5935bb399dcf23ffe24f855452b2e074bcbb3d023cfc7dd439342e54f02c6b"
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v3_core-normalized.jsonl",
+          "sha256": "2ec54da92c4e72412e89b3c07013b54729f5899ba8542eeadf962b04a1f3a4b8"
         },
-        {
-          "path": "analysis/public-roster-discovery-v1/LAYA-CORE-REUSE-VERIFIED.json",
-          "sha256": "01d2be9f23540f3fbe5d06e6707bb3d908dd76d2c509ee4af86c9e0c6129862f"
-        }
-      ]
-    },
-    "Laya-multilingual": {
-      "panels": {
         "v4": {
-          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/v4/normalized.jsonl",
-          "sha256": "814c1aa1f4d400942984ecb3786d3fd5fd15286d9b767a519c91993577795290"
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v4-normalized.jsonl",
+          "sha256": "88c576c3d62ec11686dd5ebf360971a9087779e45a346897e2badf874a754a12"
         },
         "v5": {
-          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/v5/normalized.jsonl",
-          "sha256": "6d4dbc16c7b977b15895e0e313dabc6d9db43df75ec56e89373d91f4f13ec676"
-        },
-        "old_core": {
-          "path": "eval/heldout/primary-v2/laya-multilingual/core/normalized.jsonl",
-          "sha256": "9ea39841c4d5901cd9ab860c980d2f3b55358be289c7e01b9b47982927bf849d"
-        },
-        "v3_core": {
-          "path": "eval/heldout/v3/baselines/laya-multilingual/core/normalized.jsonl",
-          "sha256": "a0b7ce9daf4be2c7831bcab6b5e96c8dcb6bef133d39475f72241ab18faccdd2"
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-4b-v5-normalized.jsonl",
+          "sha256": "d33bbb60134314dcb550060176ed0a3ae91ec0c6b033cceb40e3255b33666898"
         }
       },
       "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-baselines/laya-multilingual/ROWS.json",
-        "sha256": "d32db6cf6576f287210d31336456497f9d6926179366db29304659baf9783b43"
+        "path": "analysis/kev-reciprocal-v1/reports/kev-4b/v9-transfer-v9-test-published-temperature-rows.json",
+        "sha256": "c2cb16a936be4c10ca6b1870c69426ddd6b51629c401f71ae1da652b3bd2ac7e"
       },
       "evidence": [
         {
-          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/COMPLETE.json",
-          "sha256": "b190d683fc4c4762c2805a03d52628a53e125d228a3105c8ac2ddda00a5364e4"
-        },
-        {
-          "path": "analysis/public-roster-discovery-v1/ACTUAL-EVALUATION-COMPLETE.json",
-          "sha256": "686834a8289d3771908530a95c2ff85df71e61dbb1ce7ea95b7c9550a23ba82f"
-        },
-        {
-          "path": "analysis/decision-benchmark-v3-baselines/laya-multilingual/REPORT.json",
-          "sha256": "d20d3d58f76ca2758545f83d358abd747f86d581c897d1fb521371d1000d6154"
+          "path": "analysis/kev-reciprocal-v1/reports/kev-4b/REPORT.json",
+          "sha256": "366fb0d27a4d732e5e270c264521d9a93ccabc4f0ddf5b7d3e9f1ecc5f460406"
         },
         {
-          "path": "analysis/public-roster-discovery-v1/LAYA-CORE-REUSE-VERIFIED.json",
-          "sha256": "01d2be9f23540f3fbe5d06e6707bb3d908dd76d2c509ee4af86c9e0c6129862f"
+          "path": "analysis/kev-reciprocal-v1/our-panels/POINTS.json",
+          "sha256": "a3aa6d501a14df1cf5e7c98ecc6161c88483801310c56913465059d28c4d88ea"
         }
       ]
     },
@@ -361,99 +272,171 @@
         }
       ]
     },
-    "Lux": {
+    "Decider": {
       "panels": {
         "old_core": {
-          "path": "results/lux-new-host-v1/quality/lux/old_core/normalized.jsonl",
-          "sha256": "a659a7c7fef2871bbf806e842ea8ca81f645dcf610f0f2f6622d8d32e76a0836"
+          "path": "eval/heldout/matched-fla-v1/runs/core-fla-v1-shard1/decider/quality/normalized.jsonl",
+          "sha256": "e1b14f0f9ec521bffbde5945684bc2410a413b6b6418d65ac4fc702a7b5e7937"
         },
         "v3_core": {
-          "path": "results/lux-new-host-v1/quality/lux/v3_core/normalized.jsonl",
-          "sha256": "da8903a07e36a9687f6ee1d0fd8dbc684e61c49ef5fea30c346e46bff5a90e94"
+          "path": "eval/heldout/v3/baselines/decider/core/normalized.jsonl",
+          "sha256": "12a1f554decf1aff608743e7b4a44681289ea084ae25f383e519b1f34917d46f"
         },
         "v4": {
-          "path": "results/lux-new-host-v1/quality/lux/v4/normalized.jsonl",
-          "sha256": "26cd7482693ec1ffa1a93a156abfc88378e4872e320f3b9b775763635086d488"
+          "path": "eval/heldout/v4/baselines/decider/core/normalized.jsonl",
+          "sha256": "626ef366ae471ec10dfb89ef2ff2f7b29a6c879976d3ff14fcbb2bafd7d9b045"
         },
         "v5": {
-          "path": "results/lux-new-host-v1/quality/lux/v5/normalized.jsonl",
-          "sha256": "759c773e629a16317271de850780a8734533c795046122f850914c4482d1e72b"
+          "path": "eval/heldout/v5/open-baselines-v1/decider/core/normalized.jsonl",
+          "sha256": "ae16b17ac8e4208940c0c3e9e04f25e94686b5d16bf46f64f53669baf41aeb71"
         }
       },
       "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-baselines/Lux/ROWS.json",
-        "sha256": "bafb55fc86762061e537382360eda14a8a8f8c5e4e111c2fda1f82cd6c06793f"
+        "path": "analysis/decision-benchmark-v3-baselines/decider/ROWS.json",
+        "sha256": "6da5b5692a8a0cea3dc8f6d02d911fafd9bbf62e286a9be8db27c89f77a897e4"
       },
       "evidence": [
         {
-          "path": "results/lux-new-host-v1/quality/lux/COMPLETE.json",
-          "sha256": "9600ff792cbda13af5505c804cf63af94d2ee895f2912f0645a7d493ae1b7a9c"
+          "path": "results/accelerated-transfer-baselines-v1/decider/worker/COMPLETE.json",
+          "sha256": "e1ee0df8df59ec3eb2b49ba947a8508994a24bce226b7f2cccbe25f6835b99af"
         },
         {
-          "path": "results/lux-new-host-v1/quality/lux/metadata.json",
-          "sha256": "0d4509fae7eb3630118eaebc24e75a7b02f717d9a0ad74811fa27358c33c9cc5"
+          "path": "results/accelerated-transfer-baselines-v1/decider/worker/METADATA.json",
+          "sha256": "f00061c1b0bb642bf1353aae867d19df988c01b0428b58206b31249521bcbff8"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/Lux-projected/COMPLETE.json",
-          "sha256": "f58d02df78f297d4feebe39facfac072951111d8d6a19e2c1069f1b90250c622"
+          "path": "results/accelerated-transfer-baselines-v1/decider/worker/predictions.jsonl",
+          "sha256": "f69e87fa65b09289e6861c1d789270aea5a92d23ac47817342647ed08f406df7"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/Lux/REPORT.json",
-          "sha256": "f31703c342efe043a1e1534da189808bb7e0de0641fbaf3f91faa4a1f928a299"
+          "path": "analysis/decision-benchmark-v3-baselines/decider/REPORT.json",
+          "sha256": "5aae029f34524dfebc8e79128da8c39229c5747461dfce07ca59956da4416951"
         },
         {
-          "path": "analysis/decoder4b/lux9b-training-v1/CHECKPOINT-HELDOUT-COLLECTED.json",
-          "sha256": "d1c42515ab225b27ea011a6cf2262fe12dd7001e3b4d2cb64a23e7642c7a0d91"
+          "path": "eval/expanded-public-extension-v1/COMPARATOR-REGISTRY.json",
+          "sha256": "b80be4270721cbeeac9e28beb10c6f19f75c1cbf2553e38d3d59e3a42d56a91c"
+        },
+        {
+          "path": "analysis/decision-benchmark-v3-baselines/decider/COMPOSABLE-MANIFEST.json",
+          "sha256": "a49848e86feafaa4f36a33287910035e7605e344791aae139305b78cb304a77f"
         }
       ]
     },
-    "Decider": {
+    "Qwen3.5-4B": {
       "panels": {
         "old_core": {
-          "path": "eval/heldout/matched-fla-v1/runs/core-fla-v1-shard1/decider/quality/normalized.jsonl",
-          "sha256": "e1b14f0f9ec521bffbde5945684bc2410a413b6b6418d65ac4fc702a7b5e7937"
+          "path": "eval/heldout/primary-v2/base-4b/core/normalized.jsonl",
+          "sha256": "d20ad845066ff81caaf60ebee44df941872bdc642da78b5f3b3da4a33dd9a08d"
         },
         "v3_core": {
-          "path": "eval/heldout/v3/baselines/decider/core/normalized.jsonl",
-          "sha256": "12a1f554decf1aff608743e7b4a44681289ea084ae25f383e519b1f34917d46f"
+          "path": "eval/heldout/v3/baselines/base-4b/core/normalized.jsonl",
+          "sha256": "7541893d2b2ec0ec2ffc513f50c568eaef860b65bcef9a1747996c116fff9661"
         },
         "v4": {
-          "path": "eval/heldout/v4/baselines/decider/core/normalized.jsonl",
-          "sha256": "626ef366ae471ec10dfb89ef2ff2f7b29a6c879976d3ff14fcbb2bafd7d9b045"
+          "path": "eval/heldout/v4/baselines/base-4b/core/normalized.jsonl",
+          "sha256": "7a6d252ccaa72bf54b94445190bbad023fb465a3eaf4cf657532e2cdd4f750f5"
         },
         "v5": {
-          "path": "eval/heldout/v5/open-baselines-v1/decider/core/normalized.jsonl",
-          "sha256": "ae16b17ac8e4208940c0c3e9e04f25e94686b5d16bf46f64f53669baf41aeb71"
+          "path": "eval/heldout/v5/open-baselines-v1/base-4b/core/normalized.jsonl",
+          "sha256": "c01ad5daed65267e9aec40ed9e97ece1b6511e4cab17ebff09339b41524fb7f1"
         }
       },
       "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-baselines/decider/ROWS.json",
-        "sha256": "6da5b5692a8a0cea3dc8f6d02d911fafd9bbf62e286a9be8db27c89f77a897e4"
+        "path": "analysis/decision-benchmark-v3-baselines/base-4b/ROWS.json",
+        "sha256": "37ab17d94d640ee46283e1f23723b9ea05133c22a8a9b62a97c756a89c5de785"
       },
       "evidence": [
         {
-          "path": "results/accelerated-transfer-baselines-v1/decider/worker/COMPLETE.json",
-          "sha256": "e1ee0df8df59ec3eb2b49ba947a8508994a24bce226b7f2cccbe25f6835b99af"
+          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/COMPLETE.json",
+          "sha256": "72abc4f87a9234f95eb3b0a72a9e7d66798779d5982c757faffb610003b1b242"
         },
         {
-          "path": "results/accelerated-transfer-baselines-v1/decider/worker/METADATA.json",
-          "sha256": "f00061c1b0bb642bf1353aae867d19df988c01b0428b58206b31249521bcbff8"
+          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/METADATA.json",
+          "sha256": "871fed92cf762187015e59174a042808dbf3fc88f7791aa0ddd530ae923196ab"
         },
         {
-          "path": "results/accelerated-transfer-baselines-v1/decider/worker/predictions.jsonl",
-          "sha256": "f69e87fa65b09289e6861c1d789270aea5a92d23ac47817342647ed08f406df7"
+          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/predictions.jsonl",
+          "sha256": "8c47edda469f8914eccc2ecafa91c245752812e830cb9f29445653102522854a"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/decider/REPORT.json",
-          "sha256": "5aae029f34524dfebc8e79128da8c39229c5747461dfce07ca59956da4416951"
+          "path": "analysis/decision-benchmark-v3-baselines/base-4b/REPORT.json",
+          "sha256": "c2a0aafbe2b7c85b3c07a4b9843182a3361654f57883311155aa3b05cfc7002d"
         },
         {
           "path": "eval/expanded-public-extension-v1/COMPARATOR-REGISTRY.json",
           "sha256": "b80be4270721cbeeac9e28beb10c6f19f75c1cbf2553e38d3d59e3a42d56a91c"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/decider/COMPOSABLE-MANIFEST.json",
-          "sha256": "a49848e86feafaa4f36a33287910035e7605e344791aae139305b78cb304a77f"
+          "path": "analysis/decision-benchmark-v3-baselines/base-4b/COMPOSABLE-MANIFEST.json",
+          "sha256": "c5966aab01f3d82ee26ed56dc26ed1277f4d3cbc976ecf28d2e8d3d190fd8fc0"
+        }
+      ]
+    },
+    "Sol": {
+      "panels": {
+        "old_core": {
+          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/old_core/normalized.jsonl",
+          "sha256": "5fc0b3e49a5de576f7621839130855c178b12d942dc357aaf6b4dd9735e0a6a8"
+        },
+        "v3_core": {
+          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v3_core/normalized.jsonl",
+          "sha256": "b2c99fbd739b849a66a55929a658de4517a7ffaa4496255566e6972ba28d4f15"
+        },
+        "v4": {
+          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v4/normalized.jsonl",
+          "sha256": "3c28302750e02d4cead0aacd66a323734e773f726c4c41db658dd6461251b28b"
+        },
+        "v5": {
+          "path": "results/expanded/sol-composition-v2/candidate/job-0/worker/v5/normalized.jsonl",
+          "sha256": "6ddb0496012b596ea122afa4db4c4158e57dfb934e9ae29eb006756b652e31e1"
+        }
+      },
+      "transfer_rows": {
+        "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Sol-attempt02/v9-transfer-v9-test-published-temperature-rows.json",
+        "sha256": "b05799417cb3715de633135c58f713960b75bdb9ed30eb8912a197d33ee3d85b"
+      },
+      "evidence": [
+        {
+          "path": "results/expanded/sol-composition-v2/QUALITY-COLLECTION.json",
+          "sha256": "9b77aac87b85cbf3a74e86af4a2b9a03fd9e1ed854356a92ff54ccd0d47a16f4"
+        },
+        {
+          "path": "analysis/kev-reciprocal-v1/reports/Decision-1.0-Sol-attempt02/REPORT.json",
+          "sha256": "5d8ebb339293687cec9089f54fb96aabf3330f16111487a49489f5e52bbb6f29"
+        }
+      ]
+    },
+    "kev-0.8b": {
+      "panels": {
+        "old_core": {
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-old_core-normalized.jsonl",
+          "sha256": "b492b4512ac164f83a319f31bbc9e086cae12f87ffe2c14f307481ffb240807c"
+        },
+        "v3_core": {
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v3_core-normalized.jsonl",
+          "sha256": "70cbfc69dfc72ac0c5ae01305ec63a9f31129c115b7068c6824cd22ead013d68"
+        },
+        "v4": {
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v4-normalized.jsonl",
+          "sha256": "761dfd69acbface101f2c21fd27d2db684c5c8b8cc425890a075729ae1f7db24"
+        },
+        "v5": {
+          "path": "analysis/kev-reciprocal-v1/our-panels/kev-0.8b-v5-normalized.jsonl",
+          "sha256": "51b6fc5f826ca8db44df43885f34d857ddd5b5a66453bbdc1a83fcf488928bdb"
+        }
+      },
+      "transfer_rows": {
+        "path": "analysis/kev-reciprocal-v1/reports/kev-0.8b/v9-transfer-v9-test-published-temperature-rows.json",
+        "sha256": "d6abed88001aeac630219466f02c1aad89fcbd354b3ae533507f3c2f7fe42898"
+      },
+      "evidence": [
+        {
+          "path": "analysis/kev-reciprocal-v1/reports/kev-0.8b/REPORT.json",
+          "sha256": "6760dcbaa7d85e6c2e893e4c80ab8e66e0b84d7cfa26ae9e25909f48a05f8de4"
+        },
+        {
+          "path": "analysis/kev-reciprocal-v1/our-panels/POINTS.json",
+          "sha256": "a3aa6d501a14df1cf5e7c98ecc6161c88483801310c56913465059d28c4d88ea"
         }
       ]
     },
@@ -507,60 +490,93 @@
         }
       ]
     },
-    "Qwen3.5-4B": {
+    "Laya-base": {
       "panels": {
-        "old_core": {
-          "path": "eval/heldout/primary-v2/base-4b/core/normalized.jsonl",
-          "sha256": "d20ad845066ff81caaf60ebee44df941872bdc642da78b5f3b3da4a33dd9a08d"
-        },
-        "v3_core": {
-          "path": "eval/heldout/v3/baselines/base-4b/core/normalized.jsonl",
-          "sha256": "7541893d2b2ec0ec2ffc513f50c568eaef860b65bcef9a1747996c116fff9661"
-        },
         "v4": {
-          "path": "eval/heldout/v4/baselines/base-4b/core/normalized.jsonl",
-          "sha256": "7a6d252ccaa72bf54b94445190bbad023fb465a3eaf4cf657532e2cdd4f750f5"
+          "path": "results/public-roster-baselines-v1/laya-english/worker/v4/normalized.jsonl",
+          "sha256": "31da1d4140da69d2677172707eb301843cf8fdfb43200cba93be910321f922ba"
         },
         "v5": {
-          "path": "eval/heldout/v5/open-baselines-v1/base-4b/core/normalized.jsonl",
-          "sha256": "c01ad5daed65267e9aec40ed9e97ece1b6511e4cab17ebff09339b41524fb7f1"
+          "path": "results/public-roster-baselines-v1/laya-english/worker/v5/normalized.jsonl",
+          "sha256": "159614ceede814b578e1e025836a595c8c71a9bec4896caf6454d7bf54c4c44d"
+        },
+        "old_core": {
+          "path": "eval/heldout/primary-v2/laya-english/core/normalized.jsonl",
+          "sha256": "850643d2daeb47a6909985003736cab0f1f0e5a87631a35e27192f0733efd756"
+        },
+        "v3_core": {
+          "path": "eval/heldout/v3/baselines/laya-english/core/normalized.jsonl",
+          "sha256": "dec19caeb8074e25289e43a367392172da21f0c2ccdb403a059e2e4a2364a64b"
         }
       },
       "transfer_rows": {
-        "path": "analysis/decision-benchmark-v3-baselines/base-4b/ROWS.json",
-        "sha256": "37ab17d94d640ee46283e1f23723b9ea05133c22a8a9b62a97c756a89c5de785"
+        "path": "analysis/decision-benchmark-v3-baselines/laya-english/ROWS.json",
+        "sha256": "19138909259400facb5d3d5402d0d736f870695fe3ee46019f2e438088c57275"
       },
       "evidence": [
         {
-          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/COMPLETE.json",
-          "sha256": "72abc4f87a9234f95eb3b0a72a9e7d66798779d5982c757faffb610003b1b242"
+          "path": "results/public-roster-baselines-v1/laya-english/worker/COMPLETE.json",
+          "sha256": "cef0da797ae14cebaf1e4b9edf9b16c3cd9a412a30435d7d2517b8ba6dbe8b98"
         },
         {
-          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/METADATA.json",
-          "sha256": "871fed92cf762187015e59174a042808dbf3fc88f7791aa0ddd530ae923196ab"
+          "path": "analysis/public-roster-discovery-v1/ACTUAL-EVALUATION-COMPLETE.json",
+          "sha256": "686834a8289d3771908530a95c2ff85df71e61dbb1ce7ea95b7c9550a23ba82f"
         },
         {
-          "path": "results/remaining-transfer-baselines-v1/base-4b/worker/predictions.jsonl",
-          "sha256": "8c47edda469f8914eccc2ecafa91c245752812e830cb9f29445653102522854a"
+          "path": "analysis/decision-benchmark-v3-baselines/laya-english/REPORT.json",
+          "sha256": "cb5935bb399dcf23ffe24f855452b2e074bcbb3d023cfc7dd439342e54f02c6b"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/base-4b/REPORT.json",
-          "sha256": "c2a0aafbe2b7c85b3c07a4b9843182a3361654f57883311155aa3b05cfc7002d"
+          "path": "analysis/public-roster-discovery-v1/LAYA-CORE-REUSE-VERIFIED.json",
+          "sha256": "01d2be9f23540f3fbe5d06e6707bb3d908dd76d2c509ee4af86c9e0c6129862f"
+        }
+      ]
+    },
+    "Laya-multilingual": {
+      "panels": {
+        "v4": {
+          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/v4/normalized.jsonl",
+          "sha256": "814c1aa1f4d400942984ecb3786d3fd5fd15286d9b767a519c91993577795290"
+        },
+        "v5": {
+          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/v5/normalized.jsonl",
+          "sha256": "6d4dbc16c7b977b15895e0e313dabc6d9db43df75ec56e89373d91f4f13ec676"
+        },
+        "old_core": {
+          "path": "eval/heldout/primary-v2/laya-multilingual/core/normalized.jsonl",
+          "sha256": "9ea39841c4d5901cd9ab860c980d2f3b55358be289c7e01b9b47982927bf849d"
         },
+        "v3_core": {
+          "path": "eval/heldout/v3/baselines/laya-multilingual/core/normalized.jsonl",
+          "sha256": "a0b7ce9daf4be2c7831bcab6b5e96c8dcb6bef133d39475f72241ab18faccdd2"
+        }
+      },
+      "transfer_rows": {
+        "path": "analysis/decision-benchmark-v3-baselines/laya-multilingual/ROWS.json",
+        "sha256": "d32db6cf6576f287210d31336456497f9d6926179366db29304659baf9783b43"
+      },
+      "evidence": [
         {
-          "path": "eval/expanded-public-extension-v1/COMPARATOR-REGISTRY.json",
-          "sha256": "b80be4270721cbeeac9e28beb10c6f19f75c1cbf2553e38d3d59e3a42d56a91c"
+          "path": "results/public-roster-baselines-v1/laya-multilingual/worker/COMPLETE.json",
+          "sha256": "b190d683fc4c4762c2805a03d52628a53e125d228a3105c8ac2ddda00a5364e4"
         },
         {
-          "path": "analysis/decision-benchmark-v3-baselines/base-4b/COMPOSABLE-MANIFEST.json",
-          "sha256": "c5966aab01f3d82ee26ed56dc26ed1277f4d3cbc976ecf28d2e8d3d190fd8fc0"
+          "path": "analysis/public-roster-discovery-v1/ACTUAL-EVALUATION-COMPLETE.json",
+          "sha256": "686834a8289d3771908530a95c2ff85df71e61dbb1ce7ea95b7c9550a23ba82f"
+        },
+        {
+          "path": "analysis/decision-benchmark-v3-baselines/laya-multilingual/REPORT.json",
+          "sha256": "d20d3d58f76ca2758545f83d358abd747f86d581c897d1fb521371d1000d6154"
+        },
+        {
+          "path": "analysis/public-roster-discovery-v1/LAYA-CORE-REUSE-VERIFIED.json",
+          "sha256": "01d2be9f23540f3fbe5d06e6707bb3d908dd76d2c509ee4af86c9e0c6129862f"
         }
       ]
     }
   },
   "task_rows": 54,
   "rank_public_models": [
-    "Jev",
     "Lux",
     "Nox",
     "kev-9b",
@@ -572,7 +588,8 @@
     "kev-0.8b",
     "Qwen3.5-2B",
     "Laya-base",
-    "Laya-multilingual"
+    "Laya-multilingual",
+    "Jev"
   ],
   "internal_extra_models_not_public_rank": [
     "llm2jev-2b",
@@ -655,7 +672,5 @@
       }
     },
     "source_sha256": "14fecede5c917ea0b1b4a8f7cfb84938f984fb4a2c721acaf7b77e2cb198ec64"
-  },
-  "frozen_protocol_sha256": "ab97f652ab890eaefdd5373dd2deeb1e26588815f1f339eee4bcaa610846b0af",
-  "presentation_amendment_sha256": "f078ecb12e7f4f2baaab522d07af6695c18d9c4669a5f61623e85c93b910fd1f"
+  }
 }

model-card-example.json

@@ -65,7 +65,7 @@
   "direct_engine_exact_response": true,
   "overflow_rejected": true,
   "overflow_message": "check: 40083 tokens exceeds max_length=16384; no truncation allowed",
-  "bundle_manifest_sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830",
+  "bundle_manifest_sha256": "7d9b06bc25a75b1f131aabd280df0ef2777bf69db98574a300067ace40c61c9c",
   "runtime": {
     "actual": {
       "torch": "2.12.0+git6bbd260",

release-manifest.json

@@ -1,12 +1,12 @@
 {
   "format": "decision-public-release-v1",
-  "status": "documentation-only-assembled",
-  "bundle_manifest_sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830",
+  "status": "qualified-runtime-and-current-documents-assembled",
+  "bundle_manifest_sha256": "7d9b06bc25a75b1f131aabd280df0ef2777bf69db98574a300067ace40c61c9c",
   "readiness_sha256": "fcbb84a9bf7294ac79e2421cdd2cbfdfc46a79fd74c41ba0965778472d1cafc1",
-  "model_card_sha256": "02678d253920bc54d7af864edbee273bb649a5268ede892a6602d6b023f66800",
+  "model_card_sha256": "41dfa047d33bdc3aa9feff55928972533ba803a30b332ca380f042268fa1e848",
   "repo_id": "llm-semantic-router/Decision-1.0-Nox",
-  "assembly_script_sha256": "1697c4abaee22bb7b60259c042488178fc71847afe4810037683f874d7b8d513",
-  "original_bundle_manifest_preserved": true,
+  "assembly_script_sha256": "a8f72b8a67a64034d6183f89c945bab611056816ba0cba43670e9becb6eb78c4",
+  "original_bundle_manifest_preserved": false,
   "files_exclude_this_manifest": true,
   "files": [
     {
@@ -17,7 +17,7 @@
     {
       "file": "DIAGNOSTICS.md",
       "bytes": 5678,
-      "sha256": "f0e3a5040f9843bf1480f0383d2f01f7e9a601347d6b1545d68bf861d1f64007"
+      "sha256": "0b290793565cc5007e9ee7bc79625812c73202c7d8bfce79d8816a0e7a93a965"
     },
     {
       "file": "Dockerfile.runtime",
@@ -26,8 +26,8 @@
     },
     {
       "file": "EVALUATION.md",
-      "bytes": 3622,
-      "sha256": "0624b6aebb35e86e8456ed734027ae4590fbf24c91775e4e228fe89454780f14"
+      "bytes": 3705,
+      "sha256": "4159b7939a107e91b13e47345a192eea489f8bd870943e4b397e094b33bba905"
     },
     {
       "file": "LICENSE",
@@ -36,14 +36,19 @@
     },
     {
       "file": "MATERIALS.json",
-      "bytes": 2379,
-      "sha256": "d52c25221dab8aa6ef1a3e8c91d6651a88c662246a00db4766f99af34e5a5271"
+      "bytes": 2111,
+      "sha256": "6591d0318677c51084db78bd3bfc09bb5cacaa6c9c2d5700fc969f64b9a7ac4c"
     },
     {
       "file": "NORMALIZATION_RUNTIME.md",
       "bytes": 756,
       "sha256": "cf8e6ce1f07687a68b6adeb98e6704b8f292cb6adc7b10e4b84e8e70aa5472e5"
     },
+    {
+      "file": "NULL_DESCRIPTION_RENDERING.json",
+      "bytes": 514,
+      "sha256": "3b531cab60cba35648ab71fe4be7fd1af619f02ee337e3beed6296fdfc8b9ec8"
+    },
     {
       "file": "QUESTION-SCALING.md",
       "bytes": 1274,
@@ -56,13 +61,13 @@
     },
     {
       "file": "README.md",
-      "bytes": 4672,
-      "sha256": "02678d253920bc54d7af864edbee273bb649a5268ede892a6602d6b023f66800"
+      "bytes": 4767,
+      "sha256": "41dfa047d33bdc3aa9feff55928972533ba803a30b332ca380f042268fa1e848"
     },
     {
       "file": "RUNTIME-RELEASE.json",
-      "bytes": 1264,
-      "sha256": "97c0c58b97685f95ad57a65b378f4e4f374ff3f7a7e3f7ef37c7960131981ffe"
+      "bytes": 679,
+      "sha256": "325d7e037acab5f61dab4d56d817e7d7e2e873d9b5f0d0df587afd6c33364a46"
     },
     {
       "file": "RUNTIME.md",
@@ -76,8 +81,8 @@
     },
     {
       "file": "SENSITIVITY.md",
-      "bytes": 751,
-      "sha256": "6240f01d42aa4364ca2091f1e2d25619c21db91035348457609d0d2f9a740048"
+      "bytes": 788,
+      "sha256": "c3ea31420db4f2e5f17cc7197197a3ddd18abdad773ffb5bf1171b67136c3bc2"
     },
     {
       "file": "SERVING_OPTIMIZATION.json",
@@ -91,13 +96,18 @@
     },
     {
       "file": "TASKS.md",
-      "bytes": 9449,
-      "sha256": "ffd8e2da2bed770855e69d54c44479bd66bf27a319684954ad8b55f742706341"
+      "bytes": 9299,
+      "sha256": "2fb04bc9c17d805e2abd5e90520df25e79d85aa5296cd42074d4b156f54bf862"
     },
     {
       "file": "USAGE.md",
-      "bytes": 5435,
-      "sha256": "46ec1078d2bb8cb54a1850809a101f07a77bb54c18b92276cb89d40245dbbc8f"
+      "bytes": 5665,
+      "sha256": "b4c2f44ce120149c7cdee3749283958f34ba9caf4b6492442818c1b6a748c0d3"
+    },
+    {
+      "file": "WEIGHTING.md",
+      "bytes": 788,
+      "sha256": "c3ea31420db4f2e5f17cc7197197a3ddd18abdad773ffb5bf1171b67136c3bc2"
     },
     {
       "file": "assets/architecture.png",
@@ -236,18 +246,18 @@
     },
     {
       "file": "assets/decision-matrix.pdf",
-      "bytes": 28175,
-      "sha256": "12f963382e810eb2adb8a9fbbc3b3b7e25e75e4c13d56a04db81419e0b565033"
+      "bytes": 28045,
+      "sha256": "5e127c0cd1a6f83b772a3e66ac2e1f4ec19913ac4938371c6f190a23c50a29fe"
     },
     {
       "file": "assets/decision-matrix.png",
-      "bytes": 324709,
-      "sha256": "29c1fc0ad52b53bfc37f76788a9a824d5b767c8ed38c2474b763c1cb941182c4"
+      "bytes": 325781,
+      "sha256": "c775737aedeecadb0824092d37014a8e2cffa90af0dafcf10d584e0db823ecf8"
     },
     {
       "file": "assets/decision-matrix.svg",
       "bytes": 43846,
-      "sha256": "95e90ed75b9d942e212b621a2968be3c5387d0bf637ca681c885062c5de93456"
+      "sha256": "d99102fafe842f2a6bdd4d455f915833d5e0e09e3a22e5211fbf86e5647247ab"
     },
     {
       "file": "assets/decision-question-scaling-600px.png",
@@ -271,18 +281,18 @@
     },
     {
       "file": "assets/decision-ranking.pdf",
-      "bytes": 24211,
-      "sha256": "2c5d77b3a2420531e25a35e74389cd74968840117f90ab9f0f79e3e245101a97"
+      "bytes": 24204,
+      "sha256": "a28d4b0e8649cb7bf2db452a11d93675631af795561726aa63432b05b5888049"
     },
     {
       "file": "assets/decision-ranking.png",
-      "bytes": 214771,
-      "sha256": "a201fa8f34771901d05b3162ed33de26f38bdba6dfbcab4579b13de017cce40f"
+      "bytes": 215215,
+      "sha256": "d252593501e8faeb9c5c8cc0a9ae9a96591e7aa8696cf5394cdfec9c0ddac3b7"
     },
     {
       "file": "assets/decision-ranking.svg",
       "bytes": 14180,
-      "sha256": "39be265fee1f5e6a0b051524fadc33786c0a923a28d771a774ea3c01ace460dc"
+      "sha256": "ab7ca66a9202587dee593714e46be760cb12ac91b6459074bcbbe11c59107d9d"
     },
     {
       "file": "assets/readout.png",
@@ -316,8 +326,8 @@
     },
     {
       "file": "bundle-manifest.json",
-      "bytes": 7992,
-      "sha256": "83876db506b2d98e3e8ce7d34310b21f97bac30d4f5aef3371053798bff08830"
+      "bytes": 8171,
+      "sha256": "7d9b06bc25a75b1f131aabd280df0ef2777bf69db98574a300067ace40c61c9c"
     },
     {
       "file": "chat_template.jinja",
@@ -326,8 +336,8 @@
     },
     {
       "file": "code/decision_api.py",
-      "bytes": 10952,
-      "sha256": "273f6f10f22d5a68b8db34cfcbd35407fb43d8030f6d7cd188bcf118dd90a152"
+      "bytes": 10978,
+      "sha256": "6716bdabca3cd2f1aa447d42a6cf53ee62d7ea97984a01f16b4c09fac8aaf17e"
     },
     {
       "file": "code/decision_model.py",
@@ -361,8 +371,8 @@
     },
     {
       "file": "metrics/benchmark.json",
-      "bytes": 2276797,
-      "sha256": "40b1da632cb586a5122fa34f85ea654e2890d6619bbe707b49afe829b4e8d3ce"
+      "bytes": 2261123,
+      "sha256": "178e2e5f58da45f9fbf378ca82b49d7878fc8e9c93eee195a5ef1e7ba0f4fc6b"
     },
     {
       "file": "metrics/comparator-coverage.json",
@@ -371,8 +381,8 @@
     },
     {
       "file": "metrics/evaluation-provenance.json",
-      "bytes": 28685,
-      "sha256": "cab27fc814933826d4a280ab280877c4683524a5ea67773e4e134a00fcf8befa"
+      "bytes": 29254,
+      "sha256": "a3718fcee42a8cfe0144b01e04346c36a09ac7958154e4561327753d87d5fe8c"
     },
     {
       "file": "metrics/expanded-quality.json",
@@ -407,7 +417,7 @@
     {
       "file": "model-card-example.json",
       "bytes": 3775,
-      "sha256": "6bb34ba80660291be6c53005571a8a1d6546852bca186fff5da915019195d3a9"
+      "sha256": "1c4fc87d543722e2c3b839cbebd30c5d1370fcaf2d897f97de13079a668ac93e"
     },
     {
       "file": "pyproject.toml",
@@ -557,10 +567,10 @@
     "MATERIALS.json": "copy",
     "metrics/semantic-consistency.json": "copy"
   },
-  "scope": "13model public display only; original benchmark values and all model files unchanged.",
-  "release_tag": "v1.3.1",
-  "change_kind": "documentation-only-public-roster",
-  "previous_main_revision": "8aea799ad37bf78ea8d3be42bde29c324524889e",
-  "previous_release_manifest_sha256": "752ff577b903fb71e1012d1b0a9ccafc83336e0a104126a7ff918389c3673198",
+  "scope": "Latest13model five-panel comparison and54task diagnostics; model tensor/tokenizer/temperature identities unchanged.",
+  "release_tag": "v1.3.2",
+  "change_kind": "null-Choice-description-runtime-and-current-materials",
+  "previous_main_revision": "46505c737a45cbe2c4ac4eee38e4f94cb520e4ed",
+  "previous_release_manifest_sha256": "f537b023f52d079fdd77ecb5b1240799959209de172cc6162a6996a009ef7d03",
   "presentation_amendment_sha256": "f078ecb12e7f4f2baaab522d07af6695c18d9c4669a5f61623e85c93b910fd1f"
 }