Compact published-model comparisons; preserve science results table

Group 27 previously sourced comparisons with similarly sized published models into 10 benchmark rows. Include reference links, parameter counts, score differences, and evaluation-scope notes. Restore the original ScienceAccelBench table and section unchanged.

README.md

@@ -39,24 +39,31 @@ Weight sizes are approximate; inference also requires memory for runtime allocat
 
 Task-held-out, localized scientific-code repair on familiar codebases, with original numerical verification. Each row compares the initial model with its corresponding fine-tuned model on identical tasks. Pass rates are percentages; gains are percentage points.
 
-Each entry shows **initial model → PhAI-IDE (gain in percentage points)**, with task counts in brackets.
+| Model | Environment | Tasks | Initial model | PhAI-IDE | Gain (pp) |
+| --- | --- | ---: | ---: | ---: | ---: |
+| 4B | PLUTO-Particles-Dust | 3 | 0.00 | **33.33** | **+33.33** |
+| 9B | LAPS | 16 | 31.25 | **50.00** | **+18.75** |
+| 9B | MITgcm-biogeo | 8 | 0.00 | **12.50** | **+12.50** |
+| 9B | PLUTO-RMHD | 7 | 0.00 | **28.57** | **+28.57** |
 
-| Model | Scientific environment results |
-| --- | --- |
-| 4B | PLUTO-Particles-Dust [3]: 0.00 → **33.33** (**+33.33**) |
-| 9B | LAPS [16]: 31.25 → **50.00** (**+18.75**)<br>MITgcm-biogeo [8]: 0.00 → **12.50** (**+12.50**)<br>PLUTO-RMHD [7]: 0.00 → **28.57** (**+28.57**) |
-
-## Public benchmark performance
+## Comparison with published models
 
-Each entry shows **initial model → PhAI-IDE (gain in percentage points)**. Scores are percentages, with identical items and settings within each model pair. Gains are calculated before rounding.
+Scores (%), grouped by benchmark and model size. Each reference entry gives its published score and the **PhAI-IDE score difference in percentage points**. Reference models are approximately the same size: 3–4B, 7–9B, and 67–72B, respectively.
 
-| Model | Reasoning | Knowledge | Code |
-| --- | --- | --- | --- |
-| 4B | BBH multistep-arithmetic-two: 92.40 → **97.60** (**+5.20**) | — | HumanEvalFix-JS (docs-style repair): 35.98 → **44.51** (**+8.54**)<br>APPS introductory: 15.40 → **17.20** (**+1.80**)<br>CRUXEval output prediction†: 32.03 → **34.38** (**+2.34**) |
-| 9B | BBH word-sorting: 25.60 → **60.40** (**+34.80**)<br>AQuA-RAT: 86.61 → **88.98** (**+2.36**)<br>MATH-500: 91.40 → **92.20** (**+0.80**)<br>GSM8K†: 89.84 → **92.97** (**+3.12**) | OpenBookQA†: 33.59 → **35.16** (**+1.56**) | — |
-| 72B | — | ARC-Easy: 83.33 → **84.64** (**+1.30**)<br>ARC-Challenge: 63.05 → **64.42** (**+1.37**)<br>OpenBookQA: 48.80 → **49.20** (**+0.40**) | LiveCodeBench execution-v2: 58.46 → **60.13** (**+1.67**)<br>BIG-bench CS algorithms: 75.76 → **82.58** (**+6.82**) |
-
-† 128-item subset. ARC and OpenBookQA use normalized multiple-choice accuracy; CRUXEval and LiveCodeBench execution-v2 measure output prediction. BBH and BIG-bench entries refer to the named tasks.
+| PhAI-IDE | Benchmark | Score | Reference models: score (difference) |
+| --- | --- | ---: | --- |
+| 4B | GSM8K | **89.06†** | [Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct#instruction-tuned-models) (3B): 77.7 (**+11.36**); [Phi-3-mini](https://arxiv.org/html/2404.14219v1) (3.8B): 82.5 (**+6.56**); [Qwen2.5-3B-Instruct](https://arxiv.org/html/2412.15115v2#S5.SS2.SSS1) (3B): 86.7 (**+2.36**); [WizardMath-Llama-RL (Llama 3.2)](https://arxiv.org/html/2308.09583v2) (3B): 85.5 (**+3.56**) |
+| 4B | BBH multistep-arithmetic-two | **97.60** | [Llama-3.2-3B-Instruct](https://huggingface.co/spaces/steampunque/benchlm/blob/d45e8600172857935610426f797a4429f2f136d6/README.md) (3.21B): 53.2 (**+44.40**); [Phi-3.5-mini-8k-instruct](https://huggingface.co/spaces/steampunque/benchlm/blob/d45e8600172857935610426f797a4429f2f136d6/README.md) (3.82B): 95.6 (**+2.00**) |
+| 4B | MATH-500 | **87.50†** | [Llama-3.2-3B-Instruct](https://huggingface.co/microsoft/Phi-4-mini-reasoning/blob/661e72224c678e7360429603fd898712ee0c74b9/README.md) (3B): 44.4 (**+43.10**); [Phi-4-Mini](https://huggingface.co/microsoft/Phi-4-mini-reasoning/blob/661e72224c678e7360429603fd898712ee0c74b9/README.md) (3.8B): 71.8 (**+15.70**) |
+| 9B | GSM8K | **92.97†** | [Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct#instruction-tuned-models) (8B): 84.5 (**+8.47**); [Qwen2.5-7B-Instruct](https://arxiv.org/html/2412.15115v2#S5.SS2.SSS1) (7B): 91.6 (**+1.37**); [Gemma-2-9B-IT](https://arxiv.org/html/2412.15115v2#S5.SS2.SSS1) (9B): 76.7 (**+16.27**); [SciTulu-7B](https://arxiv.org/html/2406.07835v2#A3) (7B): 29.5 (**+63.47**) |
+| 9B | BBH word-sorting | **60.40** | [Llama-3.1-8B-Instruct](https://huggingface.co/spaces/steampunque/benchlm/blob/d45e8600172857935610426f797a4429f2f136d6/README.md) (8.03B): 51.2 (**+9.20**); [Qwen2.5-7B-Instruct](https://huggingface.co/spaces/steampunque/benchlm/blob/d45e8600172857935610426f797a4429f2f136d6/README.md) (7.62B): 15.6 (**+44.80**) |
+| 9B | MATH-500 | **92.20** | [InternLM3-8B-Instruct](https://modelscope.cn/models/Shanghai_AI_Laboratory/internlm3-8b-instruct-gptq-int4) (8B): 83 (**+9.20**); [Qwen2.5-7B-Instruct](https://modelscope.cn/models/Shanghai_AI_Laboratory/internlm3-8b-instruct-gptq-int4) (7B): 72.4 (**+19.80**); [Llama-3.1-8B-Instruct](https://modelscope.cn/models/Shanghai_AI_Laboratory/internlm3-8b-instruct-gptq-int4) (8B): 48.4 (**+43.80**) |
+| 72B | GSM8K | **93.75†** | [Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct#instruction-tuned-models) (70B): 93 (**+0.75**); [DeepSeek-LLM-67B-Chat](https://github.com/deepseek-ai/DeepSeek-LLM#3-evaluation-results) (67B): 84.1 (**+9.65**); [Qwen2-72B-Instruct](https://arxiv.org/html/2412.15115v2#S5.SS2.SSS1) (72B): 93.2 (**+0.55**); [SciTulu-70B](https://arxiv.org/html/2406.07835v2#A3) (70B): 67.5 (**+26.25**); [WizardMath-Llama-RL (Llama 2)](https://arxiv.org/html/2308.09583v2) (70B): 92.8 (**+0.95**) |
+| 72B | AQuA-RAT | **77.56** | [Llama-2-70B-Chat](https://openreview.net/pdf?id=FvfhHucpLd) (70B): 31.32 (**+46.24**) |
+| 72B | ARC-Easy | **84.64** | [Llama-2-70B](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/evaluation/more_results.md) (70B): 76.5 (**+8.14**); [DeepSeek-LLM-67B-Chat](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/evaluation/more_results.md) (67B): 81.6 (**+3.04**) |
+| 72B | ARC-Challenge | **64.42** | [Llama-2-70B](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/evaluation/more_results.md) (70B): 59.5 (**+4.92**); [DeepSeek-LLM-67B-Chat](https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/evaluation/more_results.md) (67B): 64.1 (**+0.32**) |
+
+† PhAI-IDE score on a 128-item subset. Reference scores come from the linked publications, model cards, and independent evaluation reports; evaluation settings and sample counts vary by source. Differences describe reported scores across evaluations, rather than matched-protocol head-to-head gains. BBH entries refer to the named tasks.
 
 ## Quick start