List your company

Original Tasks on extended tau-Bench Retail domain

90Pass@1ReasoningBank
80.6483.0785.587.93Apr 16, 2026

Evaluation results

MethodLinks
ReasoningBankModel=Claude-4.5-Opus2026.0490858179
PolicyBankModel=Claude-4.5-Opus2026.0489848280
ReasoningBankModel=Gemini-3-Pro2026.0487847977
No memoryModel=Claude-4.5-Opus2026.0487827875
AWMModel=Claude-4.5-Opus2026.0487848078
AWMModel=Gemini-3-Pro2026.0485827872
PolicyBankModel=Gemini-3-Pro2026.0485857975
SynapseModel=Claude-4.5-Opus2026.0485808077
No memoryModel=Gemini-3-Pro2026.0482777370
SynapseModel=Gemini-3-Pro2026.0481807570