ACR Human Audit - Unique Correct Methods: leaderboard
The ACR study uses human reviewers to assess the correctness and variety of mathematical solutions from two unmodified models, with four candidate solutions per problem. Mean unique correct primary methods per four trajectories.
Metric: Mean unique correct primary methods per four trajectories. Source: arxiv.org. Saturation forecast: Around January 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.6 35B A3B (Non-reasoning) | 1.33 | #233 (Qwen 3.6 35B A3B) |
| 2 | Gemma 4 26B A4B | 1.27 | #353 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=acr-human-audit-unique-correct-methods · How It Works · Data refreshed daily, snapshot 2026-10-29.