{"ID":2857423,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.09133","arxiv_id":"2510.09133","title":"On the Provable Performance Guarantee of Efficient Reasoning Models","abstract":"Large reasoning models (LRMs) have achieved remarkable progress in complex problem-solving tasks. Despite this success, LRMs typically suffer from high computational costs during deployment, highlighting a need for efficient inference. A practical direction of efficiency improvement is to switch the LRM between thinking and non-thinking modes dynamically. However, such approaches often introduce additional reasoning errors and lack statistical guarantees for the performance loss, which are critical for high-stakes applications. In this work, we propose Probably Approximately Correct (PAC) reasoning that controls the performance loss under the user-specified tolerance. Specifically, we construct an upper confidence bound on the performance loss and determine a threshold for switching to the non-thinking model. Theoretically, using the threshold to switch between the thinking and non-thinking modes ensures bounded performance loss in a distribution-free manner. Our comprehensive experiments on reasoning benchmarks show that the proposed method can save computational budgets and control the user-specified performance loss.","short_abstract":"Large reasoning models (LRMs) have achieved remarkable progress in complex problem-solving tasks. Despite this success, LRMs typically suffer from high computational costs during deployment, highlighting a need for efficient inference. A practical direction of efficiency improvement is to switch the LRM between thinkin...","url_abs":"https://arxiv.org/abs/2510.09133","url_pdf":"https://arxiv.org/pdf/2510.09133v2","authors":"[\"Hao Zeng\",\"Jianguo Huang\",\"Bingyi Jing\",\"Hongxin Wei\",\"Bo An\"]","published":"2025-10-10T08:33:47Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.LG\",\"math.ST\"]","methods":"[]","has_code":false}