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

SABEval

Benchmarks

Task NameDataset NameSOTA ResultTrend
Tool Invocation RefusalSABEval (D4)
Tool Invocation Rate (TIR)91.02
15
Tool Invocation RefusalSABEval (D3)
Tool Invocation Rate (TIR)85.56
15
Tool Invocation RefusalSABEval (D2)
Tool Invocation Rate (TIR)5.46
15
Tool Invocation RefusalSABEval (D1)
Tool Invocation Rate (TIR)59.46
15
Tool Invocation RefusalSABEval 1.0 (D0)
Tool Invocation Rate (TIR)2.54
15
Showing 5 of 5 rows