List your company

Text Generation and Reranking on RJUA

34.1Rouge-LEAPO
24.2226.78529.3531.915May 27, 2026

Evaluation results

MethodLinks
EAPOBackbone=Qwen2.5-7B-Instruct2026.0534.197.738.899.467.5
W-REINFORCEBackbone=Qwen2.5-7B-Instruct2026.0533.697.337.199.266.8
GRPOBackbone=Qwen2.5-7B-Instruct2026.0532.396.834.69965.7
PSRBackbone=Qwen2.5-7B-Instruct2026.0528.794.429.49762.4
NSRBackbone=Qwen2.5-7B-Instruct2026.0524.685.835.999.461.4