mlsanityData to 7 Oct 2026

RedStar: Does Scaling Long-CoT Data Unlock Better Slow-Reasoning Systems?

arXiv 2501.11284 ↗First posted 20 Jan 2025cs.LG cs.AI cs.CLcited by 60

On the map: LLM reasoning → Prompting and chain-of-thought reasoning

Abstract (arXiv)

Can scaling transform reasoning? In this work, we explore the untapped potential of scaling Long Chain-of-Thought (Long-CoT) data to 1000k samples, pioneering the development of a slow-thinking model, RedStar. Through extensive experiments with various LLMs and different sizes, we uncover the ingredients for specialization and scale for Long-CoT training. Surprisingly, even smaller models show significant performance gains with limited data, revealing the sample efficiency of Long-CoT and the critical role of sample difficulty in the learning process. Our findings demonstrate that Long-CoT reasoning can be effectively triggered with just a few thousand examples, while larger models achieve unparalleled improvements. We also introduce reinforcement learning (RL)-scale training as a promising direction for advancing slow-thinking systems. RedStar shines across domains: on the MATH-Hard benchmark, RedStar-code-math boosts performance from 66.2% to 81.6%, and on the USA Math Olympiad (AIME), it solves 46.7% of problems using only 21k mixed-code-math datasets. In multimodal tasks like GeoQA and MathVista-GEO, RedStar-Geo achieves competitive results with minimal Long-CoT data, outperforming other slow-thinking systems like QvQ-Preview. Compared to QwQ, RedStar strikes the perfect balance between reasoning and generalizability. Our work highlights that, with careful tuning, scaling Long-CoT can unlock extraordinary reasoning capabilities-even with limited dataset and set a new standard for slow-thinking models across diverse challenges. Our data and models are released at https://huggingface.co/RedStar-Reasoning.

Key facts. Cited by 60 papers in the map (data through 7 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 33 of its 61 references that are in the map. First posted on arXiv 20 Jan 2025.

Builds on

33 of 61 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them17 Oct

Other references · 32

  1. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  2. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  3. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  4. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  5. 2023Code Llama: Open Foundation Models for Codecited by 2,105
  6. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,758
  7. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,582
  8. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  9. 2017RACE: Large-scale ReAding Comprehension Dataset From Examinationscited by 851
  10. 2023Reinforced Self-Training (ReST) for Language Modelingcited by 486
  11. 2021Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoningcited by 376
  12. 2024ReST-MCTS*: LLM Self-Training via Process Reward Guided Tree Searchcited by 338
20 more
  1. 2023Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Modelscited by 325
  2. 2024Quiet-STaR: Language Models Can Teach Themselves to Think Before Speakingcited by 292
  3. 2024HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMscited by 257
  4. 2023LongLoRA: Efficient Fine-tuning of Long-Context Large Language Modelscited by 243
  5. 2024ReFT: Reasoning with Reinforced Fine-Tuningcited by 236
  6. 2021GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoningcited by 224
  7. 2024V-STaR: Training Verifiers for Self-Taught Reasonerscited by 214
  8. 2024Marco-o1: Towards Open Reasoning Models for Open-Ended Solutionscited by 142
  9. 2024Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Modelscited by 112
  10. 2024o1-Coder: an o1 Replication for Codingcited by 77
  11. 2025Virgo: A Preliminary Exploration on Reproducing o1-like MLLMcited by 68
  12. 2023A Multi-Modal Neural Geometric Solver with Textual Clauses Parsed from Diagramcited by 55
  13. 2024Scaling of Search and Learning: A Roadmap to Reproduce o1 from Reinforcement Learning Perspectivecited by 49
  14. 2024HelloBench: Evaluating Long Text Generation Capabilities of Large Language Modelscited by 42
  15. 2024CMMU: A Benchmark for Chinese Multi-modal Multi-type Question Understanding and Reasoningcited by 31
  16. 2024DRT: Deep Reasoning Translation via Long Chain-of-Thoughtcited by 28
  17. 2024GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-trainingcited by 23
  18. 2023No Train Still Gain. Unleash Mathematical Reasoning of Large Language Models with Monte Carlo Tree Search Guided by Energy Functioncited by 19
  19. 2024AtomThink: Multimodal Slow Thinking with Atomic Step Reasoningcited by 16
  20. 2024Quest: Query-centric Data Synthesis Approach for Long-context Scaling of Large Language Modelcited by 15

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 51; 2026: 9.

60 papers.

Sort:

By area:

  1. 2025s1: Simple test-time scalingcited by 1,311
  2. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  3. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 319
  4. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?cited by 215
  5. 2025Multimodal Chain-of-Thought Reasoning: A Comprehensive Surveycited by 176
  6. 2025Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyondcited by 175
  7. 2025Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretrainingcited by 123
  8. 2025A Survey of Frontiers in LLM Reasoning: Inference Scaling, Learning to Reason, and Agentic Systemscited by 108
  9. 2025Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Modelscited by 89
  10. 2025More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Modelscited by 84
  11. 2025Efficient Inference for Large Reasoning Models: A Surveycited by 58
  12. 2025Embodied-Reasoner: Synergizing Visual Search, Reasoning, and Action for Embodied Interactive Taskscited by 43
38 more
  1. 2026Uni-OPD: Unifying On-Policy Distillation with a Dual-Perspective Recipecited by 43
  2. 2025Seek in the Dark: Reasoning via Test-Time Instance-Level Policy Gradient in Latent Spacecited by 40
  3. 2025ParaThinker: Native Parallel Thinking as a New Paradigm to Scale LLM Test-time Computecited by 36
  4. 2025Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineeringcited by 34
  5. 2025VisualToolAgent (VisTA): A Reinforcement Learning Framework for Visual Tool Selectioncited by 34
  6. 2025Theoretical Physics Benchmark (TPBench) -- a Dataset and Study of AI Reasoning Capabilities in Theoretical Physicscited by 33
  7. 2025100 Days After DeepSeek-R1: A Survey on Replication Studies and More Directions for Reasoning Language Modelscited by 29
  8. 2025BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokenscited by 27
  9. 2025Genius: A Generalizable and Purely Unsupervised Self-Training Framework For Advanced Reasoningcited by 24
  10. 2026Purified OPSD: On-Policy Self-Distillation Without Losing How to Thinkcited by 23
  11. 2025A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Modelscited by 22
  12. 2025STAR-R1: Spatial TrAnsformation Reasoning by Reinforcing Multimodal LLMscited by 21
  13. 2025A Survey of Slow Thinking-based Reasoning LLMs using Reinforced Learning and Inference-time Scaling Lawcited by 18
  14. 2025Can One Domain Help Others? A Data-Centric Study on Multi-Domain Reasoning via Reinforcement Learningcited by 17
  15. 2025ThinkPatterns-21k: A Systematic Study on the Impact of Thinking Patterns in LLMscited by 16
  16. 2025TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning Compressioncited by 16
  17. 2025Why Distillation can Outperform Zero-RL: The Role of Flexible Reasoningcited by 15
  18. 2025Generative AI Act II: Test Time Scaling Drives Cognition Engineeringcited by 14
  19. 2025a1: Steep Test-time Scaling Law via Environment Augmented Generationcited by 13
  20. 2025Activation Control for Efficiently Eliciting Long Chain-of-thought Ability of Language Modelscited by 12
  21. 2025UORA: Uniform Orthogonal Reinitialization Adaptation in Parameter-Efficient Fine-Tuning of Large Modelscited by 11
  22. 2025Through the Valley: Path to Effective Long CoT Training for Small Language Modelscited by 10
  23. 2025Reinforcement Learning Teachers of Test Time Scalingcited by 10
  24. 2025Understanding and Steering the Cognitive Behaviors of Reasoning Models at Test-Timecited by 10
  25. 2025Uni-DPO: A Unified Paradigm for Dynamic Preference Optimization of LLMscited by 9
  26. 2025VisuRiddles: Fine-grained Perception is a Primary Bottleneck for Multimodal Large Language Models in Abstract Visual Reasoningcited by 7
  27. 2025Large Language Models as Computable Approximations to Solomonoff Inductioncited by 6
  28. 2025Style over Substance: Distilled Language Models Reason Via Stylistic Replicationcited by 5
  29. 2025ReFeed: Multi-dimensional Summarization Refinement with Reflective Reasoning on Feedbackcited by 4
  30. 2025From "Aha Moments" to Controllable Thinking: Toward Meta-Cognitive Reasoning in Large Reasoning Models via Decoupled Reasoning and Controlcited by 4
  31. 2025Training Multi-Image Vision Agents via End2End Reinforcement Learningcited by 4
  32. 2025Warm Up Before You Train: Unlocking General Reasoning in Resource-Constrained Settingscited by 3
  33. 2025Well Begun, Half Done: Reinforcement Learning with Prefix Optimization for LLM Reasoningcited by 3
  34. 2025Perovskite-LLM: Knowledge-Enhanced Large Language Models for Perovskite Solar Cell Researchcited by 2
  35. 2025EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensationcited by 2
  36. 2025Logit Arithmetic Elicits Long Reasoning Capabilities Without Trainingcited by 2
  37. 2025Attention-Aligned Reasoning for Large Language Modelscited by 2
  38. 2025Logit Arithmetic Elicits Long Reasoning Capabilities Without Trainingcited by 2

50 most cited of 60 All 60 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Demystifying Long Chain-of-Thought Reasoning in LLMs
  2. 2025Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training
  3. 2025Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoning
  4. 2025Why Does Reasoning Length Converge? Unveiling the Underfitting-Overfitting Trade-off in Chain-of-Thought
  5. 2025When More is Less: Understanding Chain-of-Thought Length in LLMs
  6. 2025NPG-Muse: Scaling Long Chain-of-Thought Reasoning with NP-Hard Graph Problems
  7. 2025FastCuRL: Curriculum Reinforcement Learning with Stage-wise Context Scaling for Efficient Training R1-like Reasoning Models
  8. 2025Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models