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

TinyAgent

Benchmarks

Task NameDataset NameSOTA ResultTrend
Function SelectionTinyAgent
Function Selection Accuracy94.2
50
Function CallingTinyAgent
FSA0.807
18
Tool-use Agent TasksTinyAgent 500 samples (evaluation)
Accuracy66.8
12
Tool RetrievalTinyAgent
MRR93
9
Tool Retrieval and Function SelectionTinyAgent
Function Selection Accuracy0.651
9
Showing 5 of 5 rows