List your company

Safety Evaluation on Aggregate JBB, SR, WJ

13.9Reasoning AverageIPO
11.5427.4743.459.33Sep 29, 2025

Evaluation results

MethodLinks
IPOBackbone=Qwen3-8B2025.0913.98.4
IPOBackbone=DeepSeek-R1-Distill-Llama-8B2025.0915.36.9
IPOBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0918.47.1
GRPOBackbone=DeepSeek-R1-Distill-Llama-8B2025.0918.517.8
STARBackbone=DeepSeek-R1-Distill-Llama-8B2025.0922.612.5
GRPOBackbone=Qwen3-8B2025.0923.318.7
GRPOBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0924.720
STARBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0926.516.5
SafeKeyBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0928.119.2
SafeKeyBackbone=DeepSeek-R1-Distill-Llama-8B2025.0930.918.8
RealSafeBackbone=DeepSeek-R1-Distill-Llama-8B2025.0934.22.7
RealSafeBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0936.41.7
BaseBackbone=Qwen3-8B2025.0951.334.1
SafeChainBackbone=DeepSeek-R1-Distill-Llama-8B2025.0959.439.9
SafeChainBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0960.742.2
BaseBackbone=DeepSeek-R1-Distill-Llama-8B2025.0971.556.1
BaseBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0972.960.9