MLS-Bench (Vanilla) - Structured and Causal Reasoning: leaderboard
MLS-Bench (Vanilla) Structured and Causal Reasoning scores the first proposal each model tests on the 10 Structured and Causal Reasoning tasks of MLS-Bench (2026), where a model must invent a better ML method inside a frozen research codebase. Task scores are normalized so the worst reproduced human baseline is 0 and the best is 50, on a 0 to 100 scale where higher is better.
Metric: Mean normalized task score (0-100): each metric is anchored so the worst reproduced human baseline scores 0 and the best 50 (a metric's theoretical optimum maps to 100), metrics are averaged with human weights within a setting and by geometric mean across at least three generalization settings, over the 10 MLS-Bench tasks of the Structured and Causal Reasoning area; Vanilla protocol: the score of the model's first tested proposal (tools: edit, test, submit, undo); high reasoning effort with a 10,000-token thinking budget, web search disabled, fixed seeds; higher is better. Source: arxiv.org. Saturation forecast: Around 2028. 5 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 4.6 | Thinking | 26.6 | #113 |
| 2 | DeepSeek V3.2 | Thinking | 21.6 | #419 |
| 3 | GPT-5.4 | High | 15.8 | #135 |
| 4 | Gemini 3.1 Pro (Preview) | High | 12.4 | #99 |
| 5 | Qwen 3.6 Plus | Thinking | 1.9 | #238 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mls-bench-vanilla-structured-and-causal-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.