List your company

Question Answering on HotpotQA Extreme

23.59F1 ScoreOpenDecoder
0.72046.657712.59518.5323Jan 13, 2026

Evaluation results

MethodLinks
OpenDecoderBackbone=Qwen-2.5-3B-Instruct2026.0123.5917.2
RbFTBackbone=Qwen-2.5-3B-Instruct2026.0122.1115.5
Vanilla SFTBackbone=Qwen-2.5-3B-Instruct2026.0118.2615.07
AstuteRAGBackbone=Qwen-2.5-3B-Instruct2026.0114.3810.6
InstructRAGBackbone=Qwen-2.5-3B-Instruct2026.019.145.8
Vanilla RAGBackbone=Qwen-2.5-3B-Instruct2026.014.29.67
RobustRAGBackbone=Qwen-2.5-3B-Instruct2026.011.64.67