{"ID":2851605,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.19363","arxiv_id":"2510.19363","title":"LoongRL: Reinforcement Learning for Advanced Reasoning over Long Contexts","abstract":"Reasoning over long contexts is essential for large language models. While reinforcement learning (RL) enhances short-context reasoning by inducing \"Aha\" moments in chain-of-thought, the advanced thinking patterns required for long-context reasoning remain largely unexplored, and high-difficulty RL data are scarce. In this paper, we introduce LoongRL, a data-driven RL method for advanced long-context reasoning. Central to LoongRL is KeyChain, a synthesis approach that transforms short multi-hop QA into high-difficulty long-context tasks by inserting UUID chains that hide the true question among large collections of distracting documents. Solving these tasks requires the model to trace the correct chain step-by-step, identify the true question, retrieve relevant facts and reason over them to answer correctly. RL training on KeyChain data induces an emergent plan-retrieve-reason-recheck reasoning pattern that generalizes far beyond training length. Models trained at 16K effectively solve 128K tasks without prohibitive full-length RL rollout costs. On Qwen2.5-7B and 14B, LoongRL substantially improves long-context multi-hop QA accuracy by +23.5% and +21.1% absolute gains. The resulting LoongRL-14B reaches a score of 74.2, rivaling much larger frontier models such as o3-mini (74.5) and DeepSeek-R1 (74.9). It also improves long-context retrieval, passes all 128K needle-in-a-haystack stress tests, and preserves short-context reasoning capabilities.","short_abstract":"Reasoning over long contexts is essential for large language models. While reinforcement learning (RL) enhances short-context reasoning by inducing \"Aha\" moments in chain-of-thought, the advanced thinking patterns required for long-context reasoning remain largely unexplored, and high-difficulty RL data are scarce. In...","url_abs":"https://arxiv.org/abs/2510.19363","url_pdf":"https://arxiv.org/pdf/2510.19363v2","authors":"[\"Siyuan Wang\",\"Gaokai Zhang\",\"Li Lyna Zhang\",\"Ning Shang\",\"Fan Yang\",\"Dongyao Chen\",\"Mao Yang\"]","published":"2025-10-22T08:35:28Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[\"Reinforcement Learning\",\"Language Model\"]","has_code":false}