SOURCE-LINKED INTELLIGENCE

Recognition-Refusal Misalignment in LLMs: Why Models Answer Structurally Unanswerable Questions

arXiv · AI, language, vision and robotics · article · Aug 29, 2026 · UTC

Large language models often answer structurally unanswerable questions, such as computing cot(-540°) or evaluating (1).startswith("1"), instead of abstaining. We ask whether this failure reflects missing recognition or failed routing from recognition to abstention. Across instruction-tuned models from 1.7B to 70B parameters, a single linear direction in the hidden state separates answerable from structurally impossible math and code prompts, showing that models represent impossibility before generation. Yet this recognition direction is nearly orthogonal to the canonical safety-refusal directi

Read original source ↗ Open in workspace

recordType

paper

region

Global