GR-Ben (Thinking) - Biology: leaderboard
GR-Ben (Thinking) Biology scores LLM critics with slow thinking enabled on the GR-Ben (2026) subset built from the biology problems (MMLU-Pro), where the model must name the earliest erroneous step of a step-by-step solution or state that every step is correct. The score is F1 in percent, the harmonic mean of accuracy on flawed and on correct solutions, higher is better.
Metric: F1 (%) of first-error identification: the harmonic mean of the accuracy on flawed solutions (the earliest erroneous step named) and the accuracy on fully correct solutions (judged correct), on the 399 biology problems (MMLU-Pro) of GR-Ben (3,595 human-annotated solutions written by 15 LLMs); LLMs prompted as critics with slow thinking enabled (GPT-5.2 at xhigh and Gemini 3 Flash at high reasoning effort, 32,768 output tokens); higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 7 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 3 Flash | High | 60.1 | #180 |
| 2 | GPT-5.2 (2025-12-11) | xHigh | 51.1 | |
| 3 | DeepSeek V3.2 | Thinking | 51 | #425 |
| 4 | Qwen 3 14B | 45.1 | #1284 | |
| 5 | Qwen 3 32B | 44 | #983 | |
| 6 | Qwen 3 8B | 39 | #1890 | |
| 7 | Qwen 3 4B | 35.9 | #2379 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=gr-ben-thinking-biology · How It Works · Data refreshed daily, snapshot 2026-11-22.