mlsanityData to 7 Oct 2026

Step Back to Leap Forward: Self-Backtracking for Boosting Reasoning of Language Models

arXiv 2502.04404 ↗First posted 6 Feb 2025cs.CL cs.AIcited by 27

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

Abstract (arXiv)

The integration of slow-thinking mechanisms into large language models (LLMs) offers a promising way toward achieving Level 2 AGI Reasoners, as exemplified by systems like OpenAI's o1. However, several significant challenges remain, including inefficient overthinking and an overreliance on auxiliary reward models. We point out that these limitations stem from LLMs' inability to internalize the search process, a key component of effective reasoning. A critical step toward addressing this issue is enabling LLMs to autonomously determine when and where to backtrack, a fundamental operation in traditional search algorithms. To this end, we propose a self-backtracking mechanism that equips LLMs with the ability to backtrack during both training and inference. This mechanism not only enhances reasoning ability but also efficiency by transforming slow-thinking processes into fast-thinking through self-improvement. Empirical evaluations demonstrate that our proposal significantly enhances the reasoning capabilities of LLMs, achieving a performance gain of over 40 percent compared to the optimal-path supervised fine-tuning method. We believe this study introduces a novel and promising pathway for developing more advanced and robust Reasoners.

Key facts. Cited by 27 papers in the map (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 17 of its 38 references that are in the map. First posted on arXiv 6 Feb 2025.

Builds on

17 of 38 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs30 Dec
  2. 2024Dualformer: Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Traces13 Oct
  3. 2024O1 Replication Journey: A Strategic Progress Report -- Part 18 Oct
  4. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug

Other references · 13

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  4. 2024Self-Rewarding Language Modelscited by 654
  5. 2020Generative Language Modeling for Automated Theorem Provingcited by 379
  6. 2024ReST-MCTS*: LLM Self-Training via Process Reward Guided Tree Searchcited by 338
  7. 2024KTO: Model Alignment as Prospect Theoretic Optimizationcited by 256
  8. 2022Making Large Language Models Better Reasoners with Step-Aware Verifiercited by 231
  9. 2023Learning From Mistakes Makes LLM Better Reasonercited by 97
  10. 2024Can LLMs Learn from Previous Mistakes? Investigating LLMs' Errors to Boost for Reasoningcited by 69
  11. 2024Physics of Language Models: Part 2.2, How to Learn From Mistakes on Grade-School Math Problemscited by 49
  12. 2024Backtracking Improves Generation Safetycited by 31
1 more
  1. 2024Learning to Better Search with Language Models via Guided Reinforced Self-Trainingcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 14; 2026: 13.

27 papers.

Sort:

By area:

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  2. 2025Large Language Models for Planning: A Comprehensive and Systematic Surveycited by 38
  3. 2025Where Did It All Go Wrong? A Hierarchical Look into Multi-Agent Error Attributioncited by 25
  4. 2025Reasoning-Finetuning Repurposes Latent Representations in Base Modelscited by 23
  5. 2025Auditing Meta-Cognitive Hallucinations in Reasoning Large Language Modelscited by 19
  6. 2025WebRollback: Enhancing Web Agents with Explicit Rollback Mechanismscited by 14
  7. 2025How Much Backtracking is Enough? Exploring the Interplay of SFT and RL in Enhancing LLM Reasoningcited by 13
  8. 2026NeuReasoner: Towards Explainable, Controllable, and Unified Reasoning via Mixture-of-Neuronscited by 13
  9. 2025Modeling Student Learning with 3.8 Million Program Tracescited by 10
  10. 2025First Finish Search: Efficient Test-Time Scaling in Large Language Modelscited by 8
  11. 2025Schoenfeld's Anatomy of Mathematical Reasoning by Language Modelscited by 8
  12. 2025Test-time Scaling Techniques in Theoretical Physics -- A Comparison of Methods on the TPBench Datasetcited by 7
15 more
  1. 2026Re2: Unlocking LLM Reasoning via Reinforcement Learning with Re-solvingcited by 6
  2. 2025Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoningcited by 4
  3. 2026How LLMs Detect and Correct Their Own Errors: The Role of Internal Confidence Signalscited by 3
  4. 2026STRIDE: Learnable Stepwise Language Feedback for LLM Reasoningcited by 1
  5. 2026Conformal Certification of Reasoning Trace Prefixescited by 1
  6. 2026CUSUM-Shaped Inference-Time Monitoring and Targeted Re-Decoding for Quantized Small Language Model Reasoningcited by 1
  7. 2025Adaptive Backtracking for Privacy Protection in Large Language Modelscited by 0
  8. 2025Synthetic Error Injection Fails to Elicit Self-Correction In Language Modelscited by 0
  9. 2026Caterpillar of Thoughts: The Optimal Test-Time Algorithm for Large Language Modelscited by 0
  10. 2026Behavior Cue Reasoning: Monitorable Reasoning Improves Efficiency and Safety through Oversightcited by 0
  11. 2026The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modescited by 0
  12. 2026Provable Benefits of RLVR over SFT for Reasoning Models: Learning to Backtrack Efficientlycited by 0
  13. 2026Search, Fail, Recover: A Training Framework for Correction-Aware Reasoningcited by 0
  14. 2026Calibrated e-CUSUM Decoding for Quantized Reasoning Models: Why Token Log-Probability Is the Wrong Observable for Decoding Monitorscited by 0
  15. 2026Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utilitycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026LogicTrack: Auditing Reasoning Trajectories of Large Language Models with Formal Logic Solvers
  2. 2025Advancing Reasoning in Large Language Models: Promising Methods and Approaches
  3. 2025Double-Checker: Enhancing Reasoning of Slow-Thinking LLMs via Self-Critical Fine-Tuning
  4. 2026Can We Break LLMs Out of Self-Loops? Fine-Grained Reasoning Control with Activation Steering
  5. 2026Emergent Search and Backtracking in Latent Reasoning Models
  6. 2025Dancing with Critiques: Enhancing LLM Reasoning with Stepwise Natural Language Self-Critique
  7. 2024Language Models are Hidden Reasoners: Unlocking Latent Reasoning Capabilities via Self-Rewarding
  8. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models