AI RESEARCH
LLM Agents Already Know When to Call Tools -- Even Without Reasoning
arXiv CS.CL
•
ArXi:2605.09252v2 Announce Type: replace Tool-augmented LLM agents tend to call tools indiscriminately, even when the model can answer directly. Each unnecessary call wastes API fees and latency, yet no existing benchmark systematically studies when a tool call is actually needed.