List your company

Question Answering on NQ (Held-out)

52.5AccuracyImp-SCR
41.26844.18447.150.016May 11, 2026

Evaluation results

MethodLinks
Imp-SCRBackbone=Qwen-2.5-7B, Setting=Zero-shot2026.0552.564.5——
LODO SABERBackbone=Qwen-2.5-7B, Setting=Zero-shot2026.0551.778.1-0.813.6
LODO SABERBackbone=Llama-3.1-8B, Setting=Zero-shot2026.0551.263.350.2
LODO SABERBackbone=Qwen-2.5-3B, Setting=Zero-shot2026.0548.874.673.6
LODO SABERBackbone=Llama-3.2-3B, Setting=Zero-shot2026.0547.964.54.12.8
Exp-SCRBackbone=Llama-3.1-8B, Setting=Zero-shot2026.0546.363——
Int-EvalBackbone=Llama-3.2-3B, Setting=Zero-shot2026.0543.861.8——
Int-EvalBackbone=Qwen-2.5-3B, Setting=Zero-shot2026.0541.771——