List your company

Long-context understanding on Average MuSiQue, 2WikiMQA, MFQA-En, NarrativeQA, En.QA

42.48SubEM ScoreLogo-PO
34.981636.928338.87540.8217Aug 19, 2025

Evaluation results

MethodLinks
Logo-POBackbone=Qwen2.5-14B-Instruct2025.0842.4842.46
LongMabBackbone=Qwen2.5-14B-Instruct2025.0842.1545.84
LongFaith-POBackbone=Qwen2.5-14B-Instruct2025.0840.8943.88
Vanilla LLMBackbone=Qwen2.5-14B-Instruct2025.0840.6439.57
SeaLong-POBackbone=Qwen2.5-14B-Instruct2025.0840.3337.88
LongReward-POBackbone=Qwen2.5-14B-Instruct2025.0840.2639.98
LongMab-SFTBackbone=Qwen2.5-14B-Instruct2025.0838.7541.81
LongAlpacaBackbone=Qwen2.5-14B-Instruct2025.0838.1239.98
LongAlignBackbone=Qwen2.5-14B-Instruct2025.0835.2740.4