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

Social Reasoning on ToMBench

78.34Accuracy

O3

Updated 1mo ago

Evaluation Results

Method	Links
O3 2026.03		78.34
GPT-5 2026.03		78.3
GPT-5 2026.03		78.23
GPT-4o 2026.03		74.06
O3 2026.03		74.02
Distill-Llama-70B 2026.03		73.99
GPT-4o 2026.03		73.78
Qwen3-32B 2026.03		69.75
LLaMa3.1-70B 2026.03		69.56
Qwen3-32B 2026.03		69.17
SocialR1-8B 2026.03		68.81
SocialR1-8B 2026.03		68.62
SocialR1-8B 2026.03		68.5
SocialR1-8B 2026.03		68.48
DeepSeek-R1 2026.03		68.34
SocialR1-4B 2026.03		65.74
SocialR1-4B 2026.03		64.49
SocialR1-4B 2026.03		64.49
SocialR1-8B 2026.03		64.49
SocialR1-4B 2026.03		63.76
Qwen3-4B 2026.03		63.56
Qwen3-8B 2026.03		61.79
LLaMa3.1-70B 2026.03		60.36
SocialR1-4B 2026.03		59.29
Qwen3-4B 2026.03		57.7
Qwen3-8B 2026.03		53.49