LogicScore - 2WikiMultiHopQA Essentiality: leaderboard
Metric: Essentiality (%; share of the answer's extracted propositions that lie on the recovered question-to-gold-answer path, 0 when no path is recovered; 2,177 2WikiMultiHopQA items). Source: arxiv.org. Saturation forecast: Around 2031. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 56.41 |
| 2 | Qwen 3 30B A3B | 55.99 |
| 3 | Qwen 3 14B | 54.56 |
| 4 | Gemini 3 Pro | 53.93 |
| 5 | Qwen 3 4B | 53.85 |
| 6 | Qwen 3 235B A22B | 53.25 |
| 7 | Qwen 3 8B | 52.88 |
| 8 | Qwen 3 1.7B | 52.06 |
| 9 | O3 | 50 |
| 10 | GPT-4o | 46.55 |
| 11 | DeepSeek V3.2 | 44.32 |
| 12 | GPT-5.1 | 37.96 |
| 13 | Qwen 3 0.6B | 31.53 |
Interactive version: theaggregate.ai/benchmark?slug=logicscore-2wikimultihopqa-essentiality · How It Works · Data refreshed daily, snapshot 2026-09-26.