OmniToM - Belief Extraction - Hinting Task Test: leaderboard
Metric: Belief-extraction macro F1 (%): story-level F1 between the extracted world-fact and actor-belief propositions and the gold propositions, matched semantically by a GPT-5 judge, macro-averaged over stories, on the 100 Hinting Task Test stories, on the 895 OmniToM stories from seven ToMBench categories (22,343 gold belief propositions), zero-shot TELeR level 3 prompts; GPT-5 is not scored as it is the judge; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 50.34 |
| 2 | Mistral Small 3 | 48.32 |
| 3 | Gemma 3 27B (IT) | 45.46 |
| 4 | Mistral Large 2 (Jul) | 41.78 |
| 5 | Qwen 3 32B | 41.41 |
| 6 | Qwen 3 8B | 37.14 |
| 7 | Llama 3.3 70B Instruct | 36.27 |
| 8 | Llama 3.1 8B Instruct | 31.48 |
Interactive version: theaggregate.ai/benchmark?slug=omnitom-belief-extraction-hinting-task-test · How It Works · Data refreshed daily, snapshot 2026-10-07.