BFCL

The 'BFCL' dataset/benchmark is used to evaluate the performance of tool calling in large language models by providing a structured set of tools and their interactions.

Papers using BFCL (5)

Reason Less, Verify More: Deterministic Gates Recover a Silent Policy-Violation Failure Mode in Tool-Using LLM Agents2026Looking Is Not Picking: An Attention-Segment Account of Tool-Selection Failures in LLM Agents2026ParaTool: Shifting Tool Representations from Context to Parameters2026R2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling2026GenEnv: Difficulty-Aligned Co-Evolution Between LLM Agents and Environment Simulators2025