AbstentionBench - underspecified context - BB/Disambiguate - Precision: leaderboard
Metric: Abstention Precision (%). Source: github.com. 20 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama-3.1-Tulu-3-70B | 72.73 |
| 2 | Llama-3.1-Tulu-3-70B-DPO | 63.41 |
| 3 | DeepSeek R1 Distill Llama 70B | 54.05 |
| 4 | Llama 3.1 8B | 53.57 |
| 5 | Llama-3.1-Tulu-3-8B | 53.57 |
| 6 | Llama 3.1 70B | 52.67 |
| 7 | Llama 3.1 405B Instruct | 48.53 |
| 8 | Mistral 7B Instruct (v0.3) | 47.62 |
| 9 | O1 | 46.67 |
| 10 | Qwen 2.5 32B Instruct | 46.53 |
| 11 | Llama-3.1-Tulu-3-8B-DPO | 46.15 |
| 12 | Llama 3.1 70B Instruct | 45 |
| 13 | Llama 3.3 70B Instruct | 43.98 |
| 14 | Gemini 1.5 Pro | 39.74 |
| 15 | Llama-3.1-Tulu-3-8B-SFT | 37.5 |
Interactive version: theaggregate.ai/benchmark?slug=abstentionbench-underspecified-context-bb-disambiguate-precision · How It Works · Data refreshed daily, snapshot 2026-09-19.