Residual Drift Dominates Contradiction in Multi-Turn Constraint Reasoning

Computer Science > Artificial Intelligence [Submitted on 28 Apr 2026] Residual Drift Dominates Contradiction in Multi-Turn Constraint Reasoning Sebastien Kawada How do multi-turn reasoning systems fail? The expected answer is logical contradiction, in which the system's maintained state becomes unsatisfiable. We show that the dominant mode is instead satisfiable drift, where the internal state stays consistent while the returned answer silently violates prior commitments. We build DRIFT-Bench (Decomposing Reasoning Into Failure Types), a solver-instrumented benchmark of 816 test problems across three constraint domains, and evaluate four methods on it across four open-weight models (8B-120B parameters). MUS-Repair, which feeds minimal unsatisfiable subsets back to the generator, is strongest in every setting (+1.8 to +15.0 pp over the best non-MUS baseline). But the central finding is what repair leaves behind. After structured feedback, models rarely contradict themselves. They forget. Residual errors are 98-100% satisfiable drift across all settings, while contradiction drops to near zero. Reliable multi-turn systems must separately validate that the returned answer respects the maintained state. Code is available at this https URL. Comments: Published at ICLR 2026 Workshop on Reasoning and Planning for LLMs. 18 pages. ICLR page: this https URL Code: this https URL Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) ACM classes: I.2.7 Cite as: arXiv:2605.23940 [cs.AI]   (or arXiv:2605.23940v1 [cs.AI] for this version)   https://doi.org/10.48550/arXiv.2605.23940 Focus to learn more Submission history From: Sebastien Kawada [view email] [v1] Tue, 28 Apr 2026 18:26:56 UTC (71 KB) Access Paper: HTML (experimental) view license Current browse context: cs.AI < prev   |   next > new | recent | 2026-05 Change to browse by: cs cs.CL References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)