MalruleLib: leaderboard
Metric: Malrule Reasoning Accuracy (%; predict the student's next answer from one worked mistake; cross-template, answer only). Source: arxiv.org. Saturation forecast: Around December 2026. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-OSS-120B | 56.9 |
| 2 | Qwen 3 Next 80B A3B (Thinking) | 56.4 |
| 3 | GPT-OSS-20B | 53.6 |
| 4 | Qwen 3 Next 80B A3B Instruct | 51.3 |
| 5 | Qwen 3 4B | 39.1 |
| 6 | Phi-4 | 36.7 |
| 7 | Llama 3.3 70B Instruct | 34.6 |
| 8 | Phi-4 Mini Instruct | 18.2 |
Interactive version: theaggregate.ai/benchmark?slug=malrulelib · How It Works · Data refreshed daily, snapshot 2026-09-25.