mlsanityData to 7 Oct 2026

Stepwise Reasoning Checkpoint Analysis: A Test Time Scaling Method to Enhance LLMs' Reasoning

arXiv 2505.17829 ↗First posted 23 May 2025cs.CLcited by 0

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

Abstract (arXiv)

Mathematical reasoning through Chain-of-Thought (CoT) has emerged as a powerful capability of Large Language Models (LLMs), which can be further enhanced through Test-Time Scaling (TTS) methods like Beam Search and DVTS. However, these methods, despite improving accuracy by allocating more computational resources during inference, often suffer from path homogenization and inefficient use of intermediate results. To address these limitations, we propose Stepwise Reasoning Checkpoint Analysis (SRCA), a framework that introduces checkpoints between reasoning steps. It incorporates two key strategies: (1) Answer-Clustered Search, which groups reasoning paths by their intermediate checkpoint answers to maintain diversity while ensuring quality, and (2) Checkpoint Candidate Augmentation, which leverages all intermediate answers for final decision-making. Our approach effectively reduces path homogenization and creates a fault-tolerant mechanism by utilizing high-quality intermediate results. Experimental results show that SRCA improves reasoning accuracy compared to existing TTS methods across various mathematical datasets.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 38 of its 51 references that are in the map. First posted on arXiv 23 May 2025.

Builds on

38 of 51 references.

Closest ancestors · 18

Same direction, newest first

  1. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?31 Mar
  2. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models20 Mar
  3. 2025L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning6 Mar
  4. 2025Wider or Deeper? Scaling LLM Inference-Time Compute with Adaptive Branching Tree Search6 Mar
  5. 2025Efficient Test-Time Scaling via Self-Calibration25 Feb
  6. 2025ETS: Efficient Tree Search for Inference-Time Scaling19 Feb
  7. 2025Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling10 Feb
  8. 2025Rethinking External Slow-Thinking: From Snowball Errors to Probability of Correct Reasoning26 Jan
  9. 2025A Survey of Test-Time Compute: From Intuitive Inference to Deliberate Reasoning5 Jan
  10. 2024B-STaR: Monitoring and Balancing Exploration and Exploitation in Self-Taught Reasoners23 Dec
  11. 2024Forest-of-Thought: Scaling Test-Time Compute for Enhancing LLM Reasoning12 Dec
  12. 2024Provable Scaling Laws for the Test-Time Compute of Large Language Models29 Nov
6 more
  1. 2024Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision25 Nov
  2. 2024Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions21 Nov
  3. 2024Enhancing LLM Reasoning with Reward-guided Tree Search18 Nov
  4. 2024A Comparative Study on Reasoning Patterns of OpenAI's o1 Model17 Oct
  5. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug
  6. 2024Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoning19 Jan

Other references · 20

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2025Qwen3 Technical Reportcited by 8,195
  4. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 946
  5. 2022Solving math word problems with process- and outcome-based feedbackcited by 871
  6. 2024Generative Verifiers: Reward Modeling as Next-Token Predictioncited by 465
  7. 2024Improve Mathematical Reasoning in Language Models by Automated Process Supervisioncited by 426
  8. 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinkingcited by 308
  9. 2024ProcessBench: Identifying Process Errors in Mathematical Reasoningcited by 277
  10. 2024Rewarding Progress: Scaling Automated Process Verifiers for LLM Reasoningcited by 267
  11. 2025When More is Less: Understanding Chain-of-Thought Length in LLMscited by 182
  12. 2024Accessing GPT-4 level Mathematical Olympiad Solutions via Monte Carlo Tree Self-refine with LLaMa-3 8Bcited by 156
8 more
  1. 2025Towards System 2 Reasoning in LLMs: Learning How to Think With Meta Chain-of-Thoughtcited by 108
  2. 2024OpenR: An Open Source Framework for Advanced Reasoning with Large Language Modelscited by 92
  3. 2025An Empirical Study on Eliciting and Improving R1-like Reasoning Modelscited by 78
  4. 2024Beyond Examples: High-level Automated Reasoning Paradigm in In-Context Learning via MCTScited by 52
  5. 2025Can Large Language Models Detect Errors in Long Chain-of-Thought Reasoning?cited by 43
  6. 2024Towards Self-Improvement of LLMs via MCTS: Leveraging Stepwise Knowledge with Curriculum Preference Learningcited by 27
  7. 2024Search, Verify and Feedback: Towards Next Generation Post-training Paradigm of Foundation Models via Verifier Engineeringcited by 16
  8. 2024Chain-of-Probe: Examining the Necessity and Accuracy of CoT Step-by-Stepcited by 8

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Prompting Test-Time Scaling Is A Strong LLM Reasoning Data Augmentation
  2. 2026Test-Time Scaling in Reasoning LLMs: Inference Regimes, Evaluation, and Reproducibility
  3. 2025Step-KTO: Optimizing Mathematical Reasoning through Stepwise Binary Feedback
  4. 2026ThinkBooster: A Unified Framework for Seamless Test-Time Scaling of LLM Reasoning
  5. 2025Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoning
  6. 2025From Long to Short: LLMs Excel at Trimming Own Reasoning Chains
  7. 2025ARISE: An Adaptive Resolution-Aware Metric for Test-Time Scaling Evaluation in Large Reasoning Models
  8. 2025Dynamic Early Exit in Reasoning Models