AbstentionBench - underspecified context - GSM8K - Precision: leaderboard
Metric: Abstention Precision (%). Source: github.com. 23 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.3 70B Instruct | 100 |
| 2 | GPT-4o | 100 |
| 3 | Qwen 2.5 32B Instruct | 100 |
| 4 | Llama-3.1-Tulu-3-8B | 100 |
| 5 | Llama-3.1-Tulu-3-70B | 100 |
| 6 | Llama-3.1-Tulu-3-70B-DPO | 100 |
| 7 | O1 (High) | 100 |
| 8 | O1 (Low) | 99.91 |
| 9 | O1 | 99.91 |
| 10 | Llama 3.1 70B Instruct | 99.91 |
| 11 | Gemini 1.5 Pro | 99.91 |
| 12 | Llama-3.1-Tulu-3-8B-DPO | 99.78 |
| 13 | Llama 3.1 8B Instruct | 99.73 |
| 14 | Llama 3.1 405B Instruct | 99.65 |
| 15 | Llama-3.1-Tulu-3-8B-SFT | 99.52 |
Interactive version: theaggregate.ai/benchmark?slug=abstentionbench-underspecified-context-gsm8k-precision · How It Works · Data refreshed daily, snapshot 2026-09-19.