mlsanityData to 14 Oct 2026

LoongRL: Reinforcement Learning for Advanced Reasoning over Long Contexts

arXiv 2510.19363 ↗First posted 22 Oct 2025cs.CLcited by 29

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

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.

Key facts. Cited by 29 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 32 of its 41 references that are in the map. First posted on arXiv 22 Oct 2025.

Builds on

32 of 41 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning23 May

Other references · 31

  1. 2023GPT-4 Technical Reportcited by 18,396
  2. 2025Qwen3 Technical Reportcited by 8,820
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  4. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  5. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  6. 2023Let's Verify Step by Stepcited by 3,789
  7. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,810
  8. 2024RULER: What's the Real Context Size of Your Long-Context Language Models?cited by 1,346
  9. 2024OpenAI o1 System Cardcited by 1,244
  10. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,228
  11. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,210
  12. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 1,053
19 more
  1. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 916
  2. 2024DeepSeek-V3 Technical Reportcited by 811
  3. 2023YaRN: Efficient Context Window Extension of Large Language Modelscited by 601
  4. 2025Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoningcited by 559
  5. 2024LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitaskscited by 379
  6. 2025Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRscited by 377
  7. 2024LongRoPE: Extending LLM Context Window Beyond 2 Million Tokenscited by 313
  8. 2025Reinforcement Learning for Reasoning in Large Language Models with One Training Examplecited by 295
  9. 2025MiniMax-01: Scaling Foundation Models with Lightning Attentioncited by 166
  10. 2024HELMET: How to Evaluate Long-Context Language Models Effectively and Thoroughlycited by 125
  11. 2025OpenCodeReasoning: Advancing Data Distillation for Competitive Codingcited by 113
  12. 2025rStar2-Agent: Agentic Reasoning Technical Reportcited by 54
  13. 2025rStar-Coder: Scaling Competitive Code Reasoning with a Large-Scale Verified Datasetcited by 51
  14. 2025LongReason: A Synthetic Long-Context Reasoning Benchmark via Context Expansioncited by 50
  15. 2025Search and Refine During Think: Facilitating Knowledge Refinement for Improved Retrieval-Augmented Reasoningcited by 35
  16. 2024Large Language Models Can Self-Improve in Long-context Reasoningcited by 29
  17. 2024ALR2: A Retrieve-then-Reason Framework for Long-context Question Answeringcited by 26
  18. 2025LongRoPE2: Near-Lossless LLM Context Window Scalingcited by 24
  19. 2025SelfCite: Self-Supervised Alignment for Context Attribution in Large Language Modelscited by 20

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 29.

29 papers.

Sort:

By area:

  1. 2026LongR: Unleashing Long-Context Reasoning via Reinforcement Learning with Dense Utility Rewardscited by 10
  2. 2026GoLongRL: Capability-Oriented Long Context Reinforcement Learning with Multitask Alignmentcited by 6
  3. 2026SparDA: Sparse Decoupled Attention for Efficient Long-Context LLM Inferencecited by 5
  4. 2026Beyond Teacher Likelihood: Group-Calibrated On-Policy Distillation for Long-Context Reasoningcited by 4
  5. 2026Incentivizing In-depth Reasoning over Long Contexts with Process Advantage Shapingcited by 3
  6. 2026LongAct: Harnessing Intrinsic Activation Patterns for Long-Context Reinforcement Learningcited by 3
  7. 2026Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utilitycited by 2
  8. 2026LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewardscited by 2
  9. 2026On the Generalization Gap in LLM Planning: Tests and Verifier-Reward RLcited by 1
  10. 2026A Decomposition Perspective to Long-context Reasoning for LLMscited by 1
  11. 2026How Post-Training Shapes Biological Reasoning Modelscited by 1
  12. 2026Beyond Reward Engineering: A Data Recipe for Long-Context Reinforcement Learningcited by 1
17 more
  1. 2026Evidence-State Rewards for Long-Context Reasoningcited by 1
  2. 2026Copy Less, Ground More: Overcoming Repetitive Copying in Long-Context Reasoning via Evidence-Aware Reinforcement Learningcited by 1
  3. 2026REFACT: Adaptive Fact Restatement for Compact and Faithful Chain-of-Thought Reasoningcited by 1
  4. 2026Document Reconstruction Unlocks Scalable Long-Context RLVRcited by 0
  5. 2026π2: Structure-Originated Reasoning Data Improves Long-Context Reasoning Ability of Large Language Modelscited by 0
  6. 2026Customer-Agent: Overcoming Context Limitations in Ultra-Long Shopping Trajectories via Tool-Augmented Agents and RLVRcited by 0
  7. 2026PI-Mem: Pushing Long-Context Reasoning to 3.6M Tokens with Parallel-Iterative Memorycited by 0
  8. 2026When Entropy Is Not Enough: Reclaiming Lost Semantics in LLM Output Length Predictioncited by 0
  9. 2026Not Just Reason, Not Just Scan: Reinforcement Learning for Proactive Scientific Error Verification over Academic Papercited by 0
  10. 2026Tracking the Moving Frontier: Long-Short Term Advantage Estimatorcited by 0
  11. 2026Highlight-Then-Summarize: Learning to Compress Evidence for Long-Context Understandingcited by 0
  12. 2026REALHOP: Rethinking Multi-Hop Reasoning Evaluation via Behavioral Auditingcited by 0
  13. 2026StateTree: Enhancing Long-Term Dialogue Reasoning via Reinforcement Learningcited by 0
  14. 2026LatentHarness: Learning Latent Actions for Memory and Reasoning via Counterfactual Policy Distillationcited by 0
  15. 2026TrajLong: Co-Designing Agentic and Long-Context Supervision for Mid-Trainingcited by 0
  16. 2026LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence Managementcited by 0
  17. 2026Titan-Llama: Retrofitting Pre-trained LLMs with External Neural Memory Modulescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Interleaved Reasoning for Large Language Models via Reinforcement Learning
  2. 2025Training Language Models to Reason Efficiently
  3. 2025Demystifying Long Chain-of-Thought Reasoning in LLMs
  4. 2025Rethinking Reasoning Quality in Large Language Models through Enhanced Chain-of-Thought via RL
  5. 2025ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models
  6. 2026Long-Context Reasoning Through Proxy-Based Chain-of-Thought Tuning
  7. 2025QwenLong-L1.5: Post-Training Recipe for Long-Context Reasoning and Memory Management
  8. 2025Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training