{"ID":22918765,"CreatedAt":"2026-09-17T01:02:08.507062015Z","UpdatedAt":"2026-09-20T18:11:56.143995915Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2609.17804","arxiv_id":"2609.17804","title":"A Four-Stage Decomposition of Word-Problem Solving and Mechanistic Fragility in LLM Math Reasoning","abstract":"Large language models solve grade-school math word problems with high accuracy, yet a single irrelevant clause inserted into the problem can collapse it. We reconcile these observations with a mechanistic account. We show that the model's internal computation decomposes into a four-stage sequential pipeline, Schema Abstraction, Operation Planning, Operand Binding, and Computation, each stage producing a distinct intermediate representation in an identifiable band of layers. Using the same scaffold to diagnose distractor-induced failure, we localize the corruption to a single stage, Operation Planning, implemented by a set of attention heads whose causal role we validate bidirectionally. In short, we provide a mechanistic interpretation of math word problem reasoning in LLMs, and their failure when distracted.","short_abstract":"Large language models solve grade-school math word problems with high accuracy, yet a single irrelevant clause inserted into the problem can collapse it. We reconcile these observations with a mechanistic account. We show that the model's internal computation decomposes into a four-stage sequential pipeline, Schema Abs...","url_abs":"https://arxiv.org/abs/2609.17804","url_pdf":"https://arxiv.org/pdf/2609.17804v1","authors":"[\"Zhongdi Qu\",\"Carla P. Gomes\"]","published":"2026-09-15T20:15:00Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.LG\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}