List your company

Knowledge-intensive Question Answering on TriviaQA

71.7Pass@1 AccuracyBaseline
17.131.27545.4559.625Jun 5, 2026

Evaluation results

MethodLinks
Baselinebackbone=QwQ-32B2026.0671.7630
DyConbackbone=QwQ-32B2026.0671.7592
DyConbackbone=Qwen3-14B2026.0666.8440
Baselinebackbone=Qwen3-14B2026.0666.4508
DyConbackbone=Qwen3-4B-Thinking-25072026.0634.1729
Baselinebackbone=Qwen3-4B-Thinking-25072026.0633.51,041
DyConbackbone=DeepSeek-R1-Distill-Qwen-7B2026.0619.6619
Baselinebackbone=DeepSeek-R1-Distill-Qwen-7B2026.0619.21,295