Reading Comprehension on TriviaQA RC
83.08AccuracyQwen3 30B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen3 30Bconfidence_type=normalized, evaluation_protocol=self-evaluation2026.02 | 83.08 | 0.835 | 0.124 | |
| Gemma3 12Bconfidence_type=normalized, evaluation_protocol=self-evaluation2026.02 | 80.42 | 0.831 | 0.148 | |
| GLM4 9Bconfidence_type=normalized, evaluation_protocol=self-evaluation2026.02 | 75.43 | 0.768 | 0.164 | |
| Qwen3 4Bconfidence_type=normalized, evaluation_protocol=self-evaluation2026.02 | 72.71 | 0.863 | 0.158 | |
| Gemma3 4Bconfidence_type=normalized, evaluation_protocol=self-evaluation2026.02 | 70.47 | 0.819 | 0.244 |