Document superseded repo name Compactbot/slm-architecture-benchmark-map

README.md

@@ -1,92 +1,28 @@
 ---
-license: apache-2.0
+license: mit
 language:
-  - en
+- en
 tags:
-  - small-language-model
-  - slm
-  - architecture
-  - benchmark
-  - comparison
-  - zero-shot
-  - dataset
-pipeline_tag: dataset
+- slm
+- architecture
+- benchmark
+- evaluation
+pretty_name: SLM Architecture Benchmark Scores
+size_categories:
+- n<1K
 ---
 
-# SLM Architecture → Score (controlled ablation panel)
+# SLM Architecture Benchmark Scores
 
-A small, **controlled** dataset of per-task zero-shot benchmark scores across
-different **architectures**, harvested from the model cards of the
-`d0rj/tiny-llm-ablation` family. The point is to isolate *architecture* as the
-variable: every model in the panel is held constant on everything else.
+A standardized, controlled panel of benchmark scores across small language models trained from scratch under controlled or comparable regimes.
 
-## Why this panel is controlled
+## Summary
 
-All models share:
-- **~51M parameters**, trained **from scratch** (not finetunes)
-- **Same data**: FineWeb-Edu `sample-10BT`, 3,932,160,000 source tokens
-- **Same budget**: 15,000 optimizer steps
-- **Same tokenizer**: 32,768 tokens
-- **Same eval protocol**: lm-eval 0.4.12, zero-shot, 8 tasks, full official
-  splits, 95% Wilson confidence intervals, BF16
-- **Same width/heads/FFN**: d=512, 8 query / 2 KV heads, SwiGLU ffn=1792, ctx 2048
+This dataset aggregates multi-task benchmark evaluations (ARC-Easy, BLiMP, WikiText-2, etc.) for sub-100M parameter models trained from scratch, isolating architectural differences (standard decoder-only, prefix-LM, looped/weight-tied, etc.) where experimental protocols permit.
 
-The **only** thing that varies is the architecture family. That is what makes
-an architecture→score comparison meaningful — most "which arch is best" threads
-confound architecture with scale, data and tokenizer.
-
-## The models
-
-| repo | family | what varies |
-|---|---|---|
-| `d0rj/q-51M-base` | causal-GPT | reference baseline (10 decoder layers) |
-| `d0rj/q-prefixlm-51M-base` | prefix-LM | bidirectional prefix context + suffix-only loss |
-| `d0rj/looped-51M-base` | looped (Universal-Transformer) | 10 unique blocks weight-shared × 6 loops = 60 effective layers |
-| `d0rj/diffusion-51M-base` | diffusion / masked LM | **different metric — see caveat** |
-
-> `d0rj/prefixlm-51M-base` is a **duplicate** of `q-prefixlm-51M-base` (identical
-> eval scores; the only difference is whether a 32-element RoPE buffer is counted
-> in the parameter total: 50,866,720 vs 50,866,688). It is included for
-> completeness and flagged `duplicate_of`.
-
-## Headline result (AR-comparable models only)
-
-| family | HellaSwag | ARC-E | ARC-C | PIQA | WinoG | OBQA | BoolQ | LAMBADA | **macro** |
-|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|
-| causal-GPT | 29.18 | 43.31 | 24.23 | 59.90 | 50.04 | 28.20 | 59.88 | 20.86 | **39.45** |
-| prefix-LM | 28.39 | 36.24 | 22.78 | 53.10 | 49.72 | 25.60 | 54.86 | 23.35 | **36.76** |
-| looped | 29.62 | 44.28 | 22.10 | 60.28 | 50.12 | 29.00 | 61.59 | 20.90 | **39.74** |
-
-- **Looped (weight-shared depth) ≥ causal** on 7 of 8 tasks (macro 39.74 vs 39.45); it wins most on ARC-Easy, PIQA, OBQA, BoolQ.
-- **Prefix-LM < causal** on 7 of 8 tasks (macro 36.76 vs 39.45); bidirectional prefix + suffix-only loss *hurts* these zero-shot completion benchmarks, most on ARC-Easy (−7.1) and PIQA (−6.8). Its one win is LAMBADA (+2.5), where bidirectional context helps predict the final word.
-
-## Caveats (read before trusting this)
-
-1. **n = 3 distinct AR-comparable architectures.** This is a *pairwise comparison
-   panel*, not a correlation. You cannot fit an architecture→score regression on
-   three points; the honest claim is "in this controlled panel, looped ≥ causal
-   and prefix-LM < causal", not "deeper/shared archs correlate with score".
-2. **Single training seed.** Differences of ~1–2 pts are within the 95% Wilson
-   CIs on most tasks (e.g. HellaSwag causal CI [28.30, 30.07] overlaps both
-   rivals). The directional pattern (looped up, prefix down, 7/8 tasks each) is
-   more robust than any single-task gap.
-3. **Diffusion row is a different metric.** `diffusion-51M-base` is scored with
-   continuation **perplexity** (its LAMBADA 42.21 is a PLL, not AR loglikelihood),
-   so it is **excluded from the AR macro** and must not be mixed into the
-   comparison. Its card says so explicitly.
-4. **Zero-shot, uncorrected for contamination.** One seed, no multiple-comparison
-   correction (as the source cards state).
-
-## Source & provenance
-
-Scores are author-reported `model-index` / `evaluation/results.json` values from
-the four `d0rj` repos, harvested 2026-09-24. This dataset is a *harvest +
-honest-analysis* artifact: it does not re-run the evals, it re-states the
-source numbers with the controlled-design framing and the metric caveat made
-explicit. To reproduce the underlying evals, see each repo's
-`evaluation/run_core.py`.
+Note: Consolidates and supersedes the experimental dataset name Compactbot/slm-architecture-benchmark-map.
 
 ## Files
-- `slm_arch_scores.jsonl` — one row per model: `arch` features, per-task
-  `{metric, value, ci95, n}`, `ar_macro` (null for the diffusion row),
-  `metric_type`, `duplicate_of`.
+
+- scores.csv: Tabular benchmark results and architectural metadata.
+- scores.jsonl: JSON lines format with per-model evaluation records.