ResearchBenchmarksTool Use on BFCL live and Irrelevance Detection 2025Tool Use on BFCL live and Irrelevance Detection 2025Follow88BFCL live ScoreEKSFT82.2883.76585.2586.735May 28, 2026Evaluation resultsMethodMethodLinksBFCL live ScoreIrrelevance Detection ScoreEKSFTBackbone=Qwen3-8B, Tra...Backbone=Qwen3-8B, Training Dataset=AceBench2026.0588100SFTBackbone=Qwen3-8B, Tra...Backbone=Qwen3-8B, Training Dataset=AceBench2026.0586.699.5BaseBackbone=Qwen3-8B, Tra...Backbone=Qwen3-8B, Training Dataset=AceBench2026.0584.576.9DFTBackbone=Qwen3-8B, Tra...Backbone=Qwen3-8B, Training Dataset=AceBench2026.0582.5100