{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,9,24]],"date-time":"2025-09-24T00:14:56Z","timestamp":1758672896223,"version":"3.44.0"},"publisher-location":"California","reference-count":0,"publisher":"International Joint Conferences on Artificial Intelligence Organization","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,9]]},"abstract":"<jats:p>With the rapid development of Large Language Models (LLMs), aligning these models with human preferences and values is critical to ensuring ethical and safe applications. However, existing alignment techniques such as RLHF or DPO often require direct fine-tuning on LLMs with billions of parameters, resulting in substantial computational costs and inefficiencies. To address this, we propose Micro token-level Accept-Reject Aligning (MARA) approach designed to operate independently of the language models. MARA simplifies the alignment process by decomposing sentence-level preference learning into token-level binary classification, where a compact three-layer fully-connected network determines whether candidate tokens are \u201cAccepted\u201d or \u201cRejected\u201d as part of the response. Extensive experiments across seven different LLMs and three open-source datasets show that MARA achieves significant improvements in alignment performance while reducing computational costs. The source code and implementation details are publicly available at https:\/\/github.com\/IAAR-Shanghai\/MARA, and the trained models are released at https:\/\/huggingface.co\/IAAR-Shanghai\/MARA_AGENTS.<\/jats:p>","DOI":"10.24963\/ijcai.2025\/931","type":"proceedings-article","created":{"date-parts":[[2025,9,19]],"date-time":"2025-09-19T08:10:40Z","timestamp":1758269440000},"page":"8375-8383","source":"Crossref","is-referenced-by-count":0,"title":["Token-Level Accept or Reject: A Micro Alignment Approach for Large Language Models"],"prefix":"10.24963","author":[{"given":"Yang","family":"Zhang","sequence":"first","affiliation":[{"name":"Hong Kong Polytechnic University, Hong Kong SAR, China"}]},{"given":"Yu","family":"Yu","sequence":"additional","affiliation":[{"name":"MemTensor (Shanghai) Technology Co., Ltd, Shanghai, China"}]},{"given":"Bo","family":"Tang","sequence":"additional","affiliation":[{"name":"University of Science and Technology of China, Suzhou Institute for Advanced Research, Suzhou, China"},{"name":"MemTensor (Shanghai) Technology Co., Ltd, Shanghai, China"}]},{"given":"Yu","family":"Zhu","sequence":"additional","affiliation":[{"name":"University of Science and Technology of China, Hefei, China"}]},{"given":"Chuxiong","family":"Sun","sequence":"additional","affiliation":[{"name":"China Telecom Corporation Limited Beijing Research Institute, Beijing, China"}]},{"given":"Wenqiang","family":"Wei","sequence":"additional","affiliation":[{"name":"MemTensor (Shanghai) Technology Co., Ltd, Shanghai, China"}]},{"given":"Jie","family":"Hu","sequence":"additional","affiliation":[{"name":"China Telecom Corporation Limited Beijing Research Institute, Beijing, China"}]},{"given":"Zipeng","family":"Xie","sequence":"additional","affiliation":[{"name":"Nanjing University of Information Science and Technology, Nanjing, China"}]},{"given":"Zhiyu","family":"Li","sequence":"additional","affiliation":[{"name":"MemTensor (Shanghai) Technology Co., Ltd, Shanghai, China"}]},{"given":"Feiyu","family":"Xiong","sequence":"additional","affiliation":[{"name":"MemTensor (Shanghai) Technology Co., Ltd, Shanghai, China"}]},{"given":"Edward","family":"Chung","sequence":"additional","affiliation":[{"name":"Hong Kong Polytechnic University, Hong Kong SAR, China"}]}],"member":"10584","event":{"number":"34","sponsor":["International Joint Conferences on Artificial Intelligence Organization (IJCAI)"],"acronym":"IJCAI-2025","name":"Thirty-Fourth International Joint Conference on Artificial Intelligence {IJCAI-25}","start":{"date-parts":[[2025,8,16]]},"theme":"Artificial Intelligence","location":"Montreal, Canada","end":{"date-parts":[[2025,8,22]]}},"container-title":["Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"],"original-title":[],"deposited":{"date-parts":[[2025,9,23]],"date-time":"2025-09-23T11:35:30Z","timestamp":1758627330000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ijcai.org\/proceedings\/2025\/931"}},"subtitle":[],"proceedings-subject":"Artificial Intelligence Research Articles","short-title":[],"issued":{"date-parts":[[2025,9]]},"references-count":0,"URL":"https:\/\/doi.org\/10.24963\/ijcai.2025\/931","relation":{},"subject":[],"published":{"date-parts":[[2025,9]]}}}