List your company

Commonsense Reasoning on CSQA (ACC and Length Metrics)

66.4Accuracy (CSQA)JET
43.31249.30655.361.294Sep 27, 2025

Evaluation results

MethodLinks
JETBackbone=DS-Qwen-7B2025.0966.4531
DPOBackbone=DS-Qwen-7B2025.0965.9586
AdaptThinkBackbone=DS-Qwen-7B2025.0965.9536
Laser-DBackbone=DS-Qwen-7B2025.0965.8600
DAPOBackbone=DS-Qwen-7B2025.0965.7519
SFTBackbone=DS-Qwen-7B2025.0965.4593
LCR1Backbone=DS-Qwen-7B2025.0965404
Laser-DEBackbone=DS-Qwen-7B2025.0964.8553
BaseBackbone=DS-Qwen-7B2025.0963.7631
AdaptThinkBackbone=DS-Qwen-1.5B2025.0948887
Laser-DEBackbone=DS-Qwen-1.5B2025.0947.8685
Laser-DBackbone=DS-Qwen-1.5B2025.0947.1685
SFTBackbone=DS-Qwen-1.5B2025.0947740
DAPOBackbone=DS-Qwen-1.5B2025.0945.6479
JETBackbone=DS-Qwen-1.5B2025.0945.6407
DPOBackbone=DS-Qwen-1.5B2025.0944.6690
LCR1Backbone=DS-Qwen-1.5B2025.0944.3431
BaseBackbone=DS-Qwen-1.5B2025.0944.2787