RESEARCH

Quantifying Consistency in LLM Logical Reasoning via Structural Uncertainty

ArXiv cs.AI · Wed, 17 Jun 2026 04:00:00 GMT

arXiv:2606.17312v1 Announce Type: new Abstract: Large language models can arrive at the same answer through reasoning paths that are unstable, contradictory, or difficult to rank consistently -- a failure mode especially prevalent in multi-step deductive reasoning. Existing metho

Read original source Discuss with A.S.I.S