Share your thoughts, 1 month free Claude Pro on usSee more

Lie Detection on MISID

61.54Lie Detection Accuracy

Grok-4-Fast

Updated 4d ago

Evaluation Results

Method	Links
Grok-4-Fast 2026.04		61.54	50.65
Claude-Sonnet-4.5 2026.04		61.14	57.27
GPT-5.1 2026.04		60.36	59.38
Grok-4-Fast 2026.04		59.8	45.1
Claude-Sonnet-4.5 2026.04		57.69	52.38
DeepSeek-V3 2026.04		56.6	55.05
Gemini-3-flash 2026.04		53.88	57.58
GPT-5.1 2026.04		53.85	53.15
GPT-4o 2026.04		53.85	51.61
Grok-4-Fast 2026.04		53.82	39.46
DeepSeek-V3 2026.04		50.33	49.52
DeepSeek-R1 2026.04		50.25	50.86
GPT-4o 2026.04		50.16	45.77
Gemini-2.5-flash 2026.04		50.16	36.42
GPT-4o 2026.04		49.84	48.27
Qwen3-VL-Flash 2026.04		46.15	41.58
Qwen3-VL-Max 2026.04		46.15	35.54
GLM-4.7 2026.04		42.56	45.97
Qwen3-VL-235B-A22B 2026.04		42.31	31.46
DeepSeek-R1 2026.04		42.31	43.85
Qwen3.5-Plus 2026.04		41.7	39.23
Qwen3-Max 2026.04		41.09	35.85
Qwen3-VL-Plus 2026.04		34.62	34.77
GLM-4.7 2026.04		34.62	38.27
Qwen3-Max 2026.04		34.62	29.04
Qwen3.5-Plus 2026.04		34.62	32.65
MiniMax-M2.1 2026.04		31.59	32.58
Qwen3.5-Plus 2026.04		30.77	29.65
MiniMax-M2.1 2026.04		23.08	27.31