评分 7 · 来源：cs.AI updates on arXiv.org · 发布于 2026-04-14
评分依据：Identifies structural alignment bias as a widespread flaw in LLM tool refusal - models invoke irrelevant tools due to surface-level similarity. Important finding for tool-use agent reliability.
Do LLMs Know Tool Irrelevance? Demystifying Structural Alignment Bias in Tool Invocations
发布
采集
学术前沿 7.0 分
— Identifies structural alignment bias as a widespread flaw in LLM tool refusal - models invoke irrelevant tools due to surface-level similarity. Important finding for tool-use agent reliability.