{"ID":5438854,"CreatedAt":"2026-07-01T01:17:58.482524686Z","UpdatedAt":"2026-07-03T12:27:39.719939621Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.31580","arxiv_id":"2606.31580","title":"LASER: Load-Aware Serving with Early-Exit for Reasoning LLMs at the Edge","abstract":"Large reasoning models (LRMs) such as DeepSeek-R1 have achieved strong performance through extended chain-of-thought (CoT) generation. However, deploying them on edge devices raises a conflict between long CoT sequences and constrained resources. Recent confidence-based early exit methods reduce CoT length for individual requests, yet they apply fixed thresholds from a single-request perspective, ignoring multi-request concurrency and load fluctuation in edge serving. To bridge this gap, we propose \\underline{L}oad-\\underline{A}ware \\underline{S}erving with \\underline{E}arly-exit for \\underline{R}easoning (LASER). LASER couples two complementary designs: (1) a load-aware adaptive exit threshold that adjusts the confidence bar based on real-time system load within an empirically validated robust range, and (2) a difficulty- and load-aware reasoning budget pre-allocation that assigns compute resources by request difficulty and system capacity. We formulate the problem as a joint optimization of reasoning quality and service latency. Experiments on two reasoning models, four benchmarks, and diverse load conditions show that LASER reduces average latency by 17--38\\% and improves service-level objective (SLO) satisfaction by 3--6\\% over fixed-threshold baselines, at an average accuracy cost of only 1\\%.","short_abstract":"Large reasoning models (LRMs) such as DeepSeek-R1 have achieved strong performance through extended chain-of-thought (CoT) generation. However, deploying them on edge devices raises a conflict between long CoT sequences and constrained resources. Recent confidence-based early exit methods reduce CoT length for individu...","url_abs":"https://arxiv.org/abs/2606.31580","url_pdf":"https://arxiv.org/pdf/2606.31580v1","authors":"[\"Zhiqing Tang\",\"Size Li\",\"Hanshuai Cui\",\"Zilan Huang\",\"Jianxiong Guo\",\"Tian Wang\",\"Yuan Wu\",\"Weijia Jia\"]","published":"2026-06-30T12:35:48Z","proceeding":"cs.DC","tasks":"[\"cs.DC\"]","methods":"[\"Large Language Model\"]","has_code":false}