Compliance evaluation on Compliance Evaluation Suite Average

89.99Accuracy

GPT-4o-mini (ContextLens)

Updated 4d ago

Evaluation Results

Method
GPT-4o-mini (ContextLens) 2026.04	89.99	87.89
GPT-4o (ContextLens) 2026.04	88.4	86.29
ContextReasoner-7B-PPO 2026.04	88.25	86.49
ContextReasoner-7B-SFT 2026.04	88.02	86.05
Llama-3.1-8B-Instruct (ContextLens) 2026.04	87.73	85.67
DeepSeek-R1 (671B) 2026.04	87.22	85.74
o3-mini 2026.04	86.76	85.59
Gemini-2.5-Flash (ContextLens) 2026.04	86.14	82.7
Gemini-2.5-Flash 2026.04	86.02	84.34
GPT-5 2026.04	85.79	84.09
o3-mini (ContextLens) 2026.04	85.22	82.79
o3-mini 2026.04	84.33	81.3
GPT-5 (ContextLens) 2026.04	84.09	82.1
GPT-4o 2026.04	83.76	81.71
DeepSeek-R1 (671B) (ContextLens) 2026.04	83.6	80.2
GPT-4o-mini 2026.04	82.79	81.84
GPT-4o-mini 2026.04	81.68	80.11
GPT-5 2026.04	81.2	78.2
OpenThinker-7B 2026.04	78.88	75.77
Llama-3.1-8B-Instruct 2026.04	72.78	71.71
Qwen2.5-7B-Instruct (ContextLens) 2026.04	70.61	66.73
Qwen2.5-7B-Instruct 2026.04	67.35	63.63