Publish DEVIM public research card

PUBLIC_MANIFEST.json

@@ -0,0 +1,16 @@
+{
+  "project": "DEVIM",
+  "artifact": "Evaluation methodology public frame",
+  "evaluator": "DEVIM_TR_CAPABILITY_V1_8",
+  "forced_choice_items": 2000,
+  "validator_pass": 32,
+  "validator_total": 32,
+  "hard_contamination": 0,
+  "blind_audit": {
+    "clean": 75,
+    "awkward": 5,
+    "hard": 0
+  },
+  "hidden_holdout_released": false,
+  "release_boundary": "methodology_validators_public_examples_only"
+}

README.md

@@ -0,0 +1,59 @@
+---
+language:
+- tr
+tags:
+- devim
+- evaluation
+- benchmark-methodology
+- turkish
+- research
+pretty_name: DEVİM Evaluation Research
+---
+
+# DEVİM Eval
+
+This repository documents DEVİM's evaluation methodology and evaluator failures. It does **not** release hidden holdout items.
+
+A central research lesson in DEVİM is that evaluator quality is part of the scientific problem: a model can improve language modeling while a capability instrument stays flat, and a seemingly useful QA task can turn out to be solvable by copying.
+
+## Frozen V1.8 evaluator
+
+Public-safe facts:
+
+- Evaluator: **DEVIM_TR_CAPABILITY_V1_8**
+- Forced-choice items: **2,000**
+- Validator: **32 / 32 PASS**
+- Hard contamination detected: **0**
+- Blind audit: **75 CLEAN / 5 AWKWARD / 0 HARD**
+
+The hidden item bank itself remains private to preserve evaluation integrity.
+
+## Why evaluator failures are published
+
+DEVİM's QA diagnostic exposed a copy confound: the interface could not cleanly distinguish target capability from mechanical extraction. That experiment was treated as evaluator evidence, not as a model victory.
+
+The current public principle is:
+
+**research metrics are not automatically leaderboard benchmarks.**
+
+A trustworthy public benchmark requires frozen definitions, contamination controls, validated scoring, and a release boundary that does not leak the hidden holdout.
+
+## What may be published here
+
+- evaluator methodology
+- schemas and validators
+- scoring code when safe
+- contamination-check methodology
+- public example items
+- evaluator version history
+- negative results and known confounds
+
+## What remains private
+
+- hidden V1.8 holdout items
+- sequestered evaluation material
+- private prompts/contracts
+- any artifact whose release would invalidate future measurement
+
+Project: https://devim.org  
+Source: https://github.com/bazobehram/devim-org