List your company

Physical Reasoning on I-PHYRE DeepPHY (test)

37.6Attention Score 1Qwen-3B^I GRPO
-1.32728.778918.88528.9911Mar 1, 2026

Evaluation results

MethodLinks
Qwen-3B^I GRPOPolicy Model=Qwen-3B^I GRPO, World Model=Qwen-3B^I2026.0337.651
Claude 4.0 OpusPolicy Model=Claude 4.0 Opus, World Model=-2026.0336.6756.67
Qwen-3B^I GRPOPolicy Model=Qwen-3B^I GRPO, World Model=-2026.0336.6748.89
Qwen-3B^P&I GRPOPolicy Model=Qwen-3B^P&I GRPO, World Model=-2026.0336.6760
Qwen-3B^P&I GRPOPolicy Model=Qwen-3B^P&I GRPO, World Model=Qwen-3B^P&I2026.0335.6661.99
Qwen-7BPolicy Model=Qwen-7B, World Model=-2026.0332.4232.42
Claude 4.0 OpusPolicy Model=Claude 4.0 Opus, World Model=Claude 4.0 Opus2026.033050
GPT-o3Policy Model=GPT-o3, World Model=-2026.033086.67
Qwen-7B^I GRPOPolicy Model=Qwen-7B^I GRPO, World Model=Qwen-7B^I2026.0327.8289.31
Qwen-7B^I GRPOPolicy Model=Qwen-7B^I GRPO, World Model=-2026.0326.6786.67
GPT-o3Policy Model=GPT-o3, World Model=GPT-o32026.032576.67
MOCKPolicy Model=MOCK, World Model=-2026.0323.3353.33
Qwen-3B^I SFTPolicy Model=Qwen-3B^I SFT, World Model=-2026.0323.3345.56
Qwen-3B^I SFTPolicy Model=Qwen-3B^I SFT, World Model=Qwen-3B^I2026.0323.2147.34
Qwen-7B^P&I SFTPolicy Model=Qwen-7B^P&I SFT, World Model=Qwen-7B^P&I2026.0321.4165.38
Qwen-3B^P&I SFTPolicy Model=Qwen-3B^P&I SFT, World Model=Qwen-3B^P&I2026.0321.2952.91
Gemini-2.5-ProPolicy Model=Gemini-2.5-Pro, World Model=-2026.032063.33
Qwen-3B^P SFTPolicy Model=Qwen-3B^P SFT, World Model=-2026.032030
Qwen-3B^P&I SFTPolicy Model=Qwen-3B^P&I SFT, World Model=-2026.032050
Qwen-7B^P&I SFTPolicy Model=Qwen-7B^P&I SFT, World Model=-2026.032063.33
Qwen-3B^P SFTPolicy Model=Qwen-3B^P SFT, World Model=Qwen-3B^P2026.0319.8732.45
Gemini-2.5-ProPolicy Model=Gemini-2.5-Pro, World Model=Gemini-2.5-Pro2026.0316.6753.33
Qwen-7B^P GRPOPolicy Model=Qwen-7B^P GRPO, World Model=-2026.0316.6728.89
Qwen-7B^P GRPOPolicy Model=Qwen-7B^P GRPO, World Model=Qwen-7B^P2026.0315.6830
Qwen-7B^P SFTPolicy Model=Qwen-7B^P SFT, World Model=Qwen-7B^P2026.0313.5722.15
Qwen-3BPolicy Model=Qwen-3B, World Model=-2026.0313.3316.67
Qwen-72BPolicy Model=Qwen-72B, World Model=-2026.0313.3343.33
Qwen-3B^P GRPOPolicy Model=Qwen-3B^P GRPO, World Model=-2026.0313.3326.67
Qwen-7B^P SFTPolicy Model=Qwen-7B^P SFT, World Model=-2026.0313.3319.33
Qwen-7B^P&I GRPOPolicy Model=Qwen-7B^P&I GRPO, World Model=-2026.0313.3390
Qwen-7B^P&I GRPOPolicy Model=Qwen-7B^P&I GRPO, World Model=Qwen-7B^P&I2026.0313.1293.33
Qwen-3B^P GRPOPolicy Model=Qwen-3B^P GRPO, World Model=Qwen-3B^P2026.0312.7828
Qwen-7B^I SFTPolicy Model=Qwen-7B^I SFT, World Model=-2026.0312.2250
Qwen-7B^I SFTPolicy Model=Qwen-7B^I SFT, World Model=Qwen-7B^I2026.0311.3451.5
Qwen-72BPolicy Model=Qwen-72B, World Model=Qwen-72B2026.031036.67
Qwen-32BPolicy Model=Qwen-32B, World Model=-2026.030.175.6