List your company

Tool-use Search Evaluation on SLATE synthetic

68.5Tool Match RateEGB-Sampling (Ours)
61.2263.116566.89Apr 13, 2026

Evaluation results

MethodLinks
EGB-Sampling (Ours)Model=Claude-Sonnet-42026.0468.55487.9
ReActModel=Claude-Sonnet-42026.0466.429.385.7
Baseline-LLMModel=Claude-Sonnet-42026.0465.2——
LATSModel=Claude-Sonnet-42026.0463.436.587.3
ReflexionModel=Claude-Sonnet-42026.0461.544.783.4