List your company
ResearchFormerly WizWand

Social Interaction on SOTOPIA Hard

7.93GOAL ScoreSAVOIR
5.02845.78176.5357.2883Apr 21, 2026Apr 25, 2026Apr 29, 2026May 3, 2026May 7, 2026May 11, 2026May 15, 2026
Updated 5mo ago

Evaluation results

MethodLinks
SAVOIREvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.047.933.97——
Sotopia-RLEvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.047.813.8——
DSIEvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.047.313.51——
SAVOIREvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.047.183.51——
ALSO2026.057.113.532.435.47
GPT-4oEvaluation Setting=Self-Play2026.046.973.46——
GPT-4oEvaluation Setting=GPT-4o-as-Partner2026.046.973.46——
INSTINCT2026.056.923.432.165.44
DSIEvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.046.873.42——
EPOEvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.046.823.12——
EPOEvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.046.813.51——
DATEvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.046.783.36——
EvoPrompt2026.056.773.291.935.15
PPDPPEvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.046.763.35——
OPRO2026.056.713.241.894.63
Gemini-2.5-ProEvaluation Setting=GPT-4o-as-Partner2026.046.73.09——
DeepSeek-V3Evaluation Setting=GPT-4o-as-Partner2026.046.693.31——
Sotopia-RLEvaluation Setting=GPT-4o-as-Partner, Base Model=Qwen2.5-7B-Instruct2026.046.683.29——
OpenAI-o1Evaluation Setting=GPT-4o-as-Partner2026.046.653.2——
Claude-3.5-SonnetEvaluation Setting=GPT-4o-as-Partner2026.046.643.3——
PPDPPEvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.046.633.31——
Vanilla2026.056.523.021.324.37
DATEvaluation Setting=Self-Play, Base Model=Qwen2.5-7B-Instruct2026.046.393.1——
DeepSeek-V3Evaluation Setting=Self-Play2026.046.343.09——
Claude-3.5-SonnetEvaluation Setting=Self-Play2026.046.333.09——
OpenAI-o3-miniEvaluation Setting=GPT-4o-as-Partner2026.046.332.98——
Qwen2.5-7B-InstructEvaluation Setting=Self-Play2026.046.213.01——
DeepSeek-R1Evaluation Setting=GPT-4o-as-Partner2026.046.22.95——
QwQ-32BEvaluation Setting=GPT-4o-as-Partner2026.046.192.91——
Qwen2.5-7B-InstructEvaluation Setting=GPT-4o-as-Partner2026.045.92.9——
DeepSeek-R1Evaluation Setting=Self-Play2026.045.862.73——
OpenAI-o1Evaluation Setting=Self-Play2026.045.692.71——
Gemini-2.5-ProEvaluation Setting=Self-Play2026.045.672.55——
QwQ-32BEvaluation Setting=Self-Play2026.045.352.41——
OpenAI-o3-miniEvaluation Setting=Self-Play2026.045.142.36——