RuleWorld: leaderboard
Metric: Exact-match accuracy (%; x100 of the 0-1 score; mean of the six task-representation cells (single-rule, parallel multi-rule over 2-8 rules, and 2-4 hop multi-hop QA, each with rules in first-order logic and in natural language); all 1,000 abstract, non-commonsense procedural rules of the pool prepended to the prompt (full-context prompting); exact match on the boxed answer list, partial credit per matching position; 10 questions per task type and difficulty level for each of five seeds; temperature 0 (Qwen3-32B 0.6)). Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 77.63 |
| 2 | GPT-5.5 | 72.83 |
| 3 | GPT-5.1 | 67.96 |
| 4 | DeepSeek V3.2 (Non-reasoning) | 66.87 |
| 5 | Llama 3.1 70B Instruct | 54.9 |
| 6 | Qwen 2.5 72B Instruct | 53.07 |
| 7 | Llama 3.3 70B Instruct | 51.62 |
| 8 | Qwen 3 32B | 51.32 |
| 9 | Qwen 2.5 32B Instruct | 50.77 |
| 10 | Qwen 2.5 14B Instruct | 42.97 |
Interactive version: theaggregate.ai/benchmark?slug=ruleworld · How It Works · Data refreshed daily, snapshot 2026-09-26.