{"type":"ScholarlyArticle","title":"System 1&2 Synergy via Dynamic Model Interpolation","authors":[{"name":"Chenxu Yang","slug":"chenxu-yang","given":"Chenxu","family":"Yang","affiliation":""},{"name":"Qingyi Si","slug":"qingyi-si","given":"Qingyi","family":"Si","affiliation":""},{"name":"Chong Tian","slug":"chong-tian","given":"Chong","family":"Tian","affiliation":""},{"name":"Xiyu Liu","slug":"xiyu-liu","given":"Xiyu","family":"Liu","affiliation":""},{"name":"Dingyu Yao","slug":"dingyu-yao","given":"Dingyu","family":"Yao","affiliation":""},{"name":"Chuanyu Qin","slug":"chuanyu-qin","given":"Chuanyu","family":"Qin","affiliation":""},{"name":"Zheng Lin","slug":"zheng-lin","given":"Zheng","family":"Lin","affiliation":""},{"name":"Weiping Wang","slug":"weiping-wang","given":"Weiping","family":"Wang","affiliation":""},{"name":"Jiaqi Wang","slug":"jiaqi-wang","given":"Jiaqi","family":"Wang","affiliation":""}],"abstract":"Training a unified language model that adapts between intuitive System 1 and deliberative System 2 remains challenging due to interference between their cognitive modes. Recent studies have thus pursued making System 2 models more efficient. However, these approaches focused on output control, limiting what models produce. We argue that this paradigm is misaligned: output length is merely a symptom of the model's cognitive configuration, not the root cause. In this work, we shift the focus to capability control, which modulates how models think rather than what they produce. To realize this, we leverage existing Instruct and Thinking checkpoints through dynamic parameter interpolation, without additional training. Our pilot study establishes that linear interpolation yields a convex, monotonic Pareto frontier, underpinned by representation continuity and structural connectivity. Building on this, we propose DAMI (DynAmic Model Interpolation), a framework that estimates a query-specific Reasoning Intensity lambda(q) to configure cognitive depth. For training-based estimation, we develop a preference learning method encoding accuracy and efficiency criteria. For zero-shot deployment, we introduce a confidence-based method leveraging inter-model cognitive discrepancy. Experiments on five mathematical reasoning benchmarks demonstrate that DAMI achieves higher accuracy than the Thinking model while remaining efficient, effectively combining the efficiency of System 1 with the reasoning depth of System 2.","doi":"10.24963/ijcai.2026/668","url":"https://ijcai.org/proceedings/2026/668","pdf":"https://ijcai.org/proceedings/2026/0668.pdf","markdown":"https://ijcai.org/proceedings/2026/668.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/668","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Natural Language Processing","track":"Main Track","number":668,"pages":{"first":6000,"last":6008},"keywords":["Natural Language Processing: Applications","Natural Language Processing: Language models","Natural Language Processing: Language generation"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}