List your company

Question Answering on popQA Noisy

25.07F1 ScoreOpenDecoder
7.140411.795216.4521.1048Jan 13, 2026

Evaluation results

MethodLinks
OpenDecoderBackbone=Qwen-2.5-3B-Instruct2026.0125.0722.02
RbFTBackbone=Qwen-2.5-3B-Instruct2026.0123.7120.2
Vanilla SFTBackbone=Qwen-2.5-3B-Instruct2026.0121.0618.16
AstuteRAGBackbone=Qwen-2.5-3B-Instruct2026.0120.2816.6
InstructRAGBackbone=Qwen-2.5-3B-Instruct2026.0114.2521.4
RobustRAGBackbone=Qwen-2.5-3B-Instruct2026.0110.3314.8
Vanilla RAGBackbone=Qwen-2.5-3B-Instruct2026.017.8320.66