SCOPE Experimental Design - Datasets (CoT-only): leaderboard
SCOPE (2026) asks LLMs to design the experiments of recent ICML, NeurIPS and ICLR papers from their task description, related work and method, and a GPT-5.2 judge scores the plan against the real design with a redline rule that zeros fatal flaws. Sub-dimension score out of 5 for the choice of datasets, higher is better, with chain-of-thought prompting and no search.
Metric: Sub-dimension score (out of 5) for dataset selection: rubric scores from a GPT-5.2 judge against the ground-truth experimental design of 300 recent ICML, NeurIPS and ICLR papers in 19 research domains; a redline rule zeros a sub-dimension with a fatal flaw such as a hallucinated dataset or baseline, an incompatible metric or a violated constraint; chain-of-thought prompting without search. Source: arxiv.org. Saturation forecast: Around September 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 2.18 | #172 |
| 2 | Claude Sonnet 4.5 | 2.01 | #244 |
| 3 | Gemini 3 Pro | 1.85 | #135 |
| 4 | Grok 4 | 1.8 | #306 |
| 5 | DeepSeek V3.2 | 1.76 | #403 |
| 6 | Kimi K2 | 1.68 | #435 |
| 7 | Qwen 3 Max | 1.64 | #401 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=scope-experimental-design-datasets-cot-only · How It Works · Data refreshed daily, snapshot 2026-11-12.