List your company

Question Answering on ConFiQA-QA (Held-out)

67AccuracyImp-SCR
49.11253.75658.463.044May 11, 2026

Evaluation results

MethodLinks
Imp-SCRBackbone=Qwen-2.5-7B, Setting=Zero-shot2026.056786——
Imp-SCRBackbone=Llama-3.1-8B, Setting=Zero-shot2026.0566.873.7——
LODO SABERBackbone=Llama-3.1-8B, Setting=Zero-shot2026.0565.572.9-1.3-0.8
Imp-SCRBackbone=Qwen-2.5-3B, Setting=Zero-shot2026.0557.266.9——
LODO SABERBackbone=Llama-3.2-3B, Setting=Zero-shot2026.0556.172.93.51.5
LODO SABERBackbone=Qwen-2.5-7B, Setting=Zero-shot2026.0555.871.5-11.2-14.5
Ctx-EvalBackbone=Llama-3.2-3B, Setting=Zero-shot2026.0552.671.4——
LODO SABERBackbone=Qwen-2.5-3B, Setting=Zero-shot2026.0549.865.3-7.5-1.6