{"slug": "mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization", "title": "Mitigating Factual Hallucination in LRMs via Mixed-Mode Advantage Regularization", "summary": "Researchers at an undisclosed institution introduced MARGO, a reinforcement learning framework that mitigates thinking-induced hallucination in large reasoning models by using non-thinking rollouts as references. The method improves factual reliability on question-answering benchmarks while preserving general reasoning ability.", "body_md": "# Computer Science > Computation and Language\n\n[Submitted on 7 Jul 2026]\n\n# Title:Mitigating Factual Hallucination in Large Reasoning Models via Mixed-Mode Advantage Regularization\n\n[View PDF](/pdf/2607.05861)\n\nAbstract:Large reasoning models (LRMs) improve language model capabilities by generating explicit thinking traces before final answers. In factuality-oriented question answering (QA), such thinking often improves overall performance by helping the model recover relevant knowledge and refine its answers. However, we find that this benefit is not uniform at the instance level: explicit thinking can also overturn correct non-thinking answers and lead to factual drift. We refer to this failure mode as \\emph{thinking-induced hallucination}. To explain this phenomenon, we formulate explicit thinking in factuality QA as a thinking residual over the model's direct-answer tendency, which can either recover missing knowledge or introduce unsupported associations. Based on this formulation, we propose MARGO, \\underline{\\textit{M}}ixed-Mode \\underline{\\textit{A}}dvantage \\underline{\\textit{R}}egularization for \\underline{\\textit{G}}rounded \\underline{\\textit{O}}ptimization, a reinforcement learning framework that uses non-thinking rollouts as same-model references in advantage estimation. By constructing mixed-mode rollout groups with both thinking and non-thinking trajectories, MARGO evaluates whether explicit thinking adds factual value beyond direct answering, thereby suppressing hallucination-prone thinking while preserving beneficial thinking behaviors. Experiments across multiple factuality-oriented QA benchmarks demonstrate that MARGO improves factual reliability over strong baselines, while evaluations on mathematical benchmarks show that it preserves general reasoning ability.\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer\n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers\n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps\n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations\n\n*(*[What are Smart Citations?](https://www.scite.ai/))# Code, Data and Media Associated with this Article\n\nalphaXiv\n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers\n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub\n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub\n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face\n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast\n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower\n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender\n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [ Learn more about arXivLabs](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization", "canonical_source": "https://arxiv.org/abs/2607.05861", "published_at": "2026-07-12 17:05:52+00:00", "updated_at": "2026-07-12 17:35:31.592174+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "ai-safety", "natural-language-processing"], "entities": ["MARGO"], "alternates": {"html": "https://wpnews.pro/news/mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization", "markdown": "https://wpnews.pro/news/mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization.md", "text": "https://wpnews.pro/news/mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization.txt", "jsonld": "https://wpnews.pro/news/mitigating-factual-hallucination-in-lrms-via-mixed-mode-advantage-regularization.jsonld"}}