BFCL

The BFCL dataset/benchmark is used to evaluate the performance of reasoning-aware reinforcement learning frameworks for interpretable function calling in large language models, focusing on aspects like function-calling accuracy and interpretability.

Papers using BFCL (4)

How Many Tools Should an LLM Agent See? A Chance-Corrected Answer2026R2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling2026CuES: A Curiosity-driven and Environment-grounded Synthesis Framework for Agentic RL2025Agent Safety Alignment via Reinforcement Learning2025