List your company

Clinical Reasoning on MEDREASON

73.4Pass@1Outcome Reward
23.27236.28649.362.314Oct 2, 2025

Evaluation results

MethodLinks
Outcome RewardBackbone=Llama3.1-8B2025.1073.4
Ours (Sparse)Backbone=Llama3.1-8B2025.1073.1
Ours (Interval)Backbone=Llama3.1-8B2025.1071.4
Ours (Dense)Backbone=Llama3.1-8B2025.1069.8
Ours (Sparse)Backbone=Qwen3-4B2025.1066.1
Outcome RewardBackbone=Qwen2.5-7B2025.1065
Ours (Interval)Backbone=Qwen2.5-7B2025.1063.8
SFTBackbone=Llama3.1-8B2025.1063.5
Ours (Dense)Backbone=Qwen2.5-7B2025.1058.3
SFTBackbone=Qwen3-4B2025.1057.6
Outcome RewardBackbone=Qwen3-4B2025.1056.3
Ours (Dense)Backbone=Qwen3-4B2025.1053.7
SFTBackbone=Qwen2.5-7B2025.1053
Ours (Interval)Backbone=Qwen3-4B2025.1052.1
Ours (Sparse)Backbone=Qwen2.5-7B2025.1025.2