Enhancing Long Chain-of-Thought Reasoning through Multi-Path Plan Aggregation

• URL: http://arxiv.org/abs/2510.11620v2
• Date: Thu, 16 Oct 2025 21:41:15 GMT
• Title: Enhancing Long Chain-of-Thought Reasoning through Multi-Path Plan Aggregation
• Authors: Siheng Xiong, Ali Payani, Faramarz Fekri,
• Abstract summary: We analyze raw long CoTs and uncover a reasoning hierarchy consisting of planning and execution steps.<n>Motivated by this observation, we propose Multi-Path Plan Aggregation (MPPA), a framework that augments single-pass reasoning with plan exploration and aggregation.<n>To overcome this, we introduce online Step-DPO, a process-level preference optimization scheme that leverages Twisted Sequential Monte Carlo (TSMC) to provide scalable stepwise supervision.
• Score: 32.86351316550696
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Inference-time scaling enhances the reasoning ability of a language model (LM) by extending its chain-of-thought (CoT). However, existing approaches typically generate the entire reasoning chain in a single forward pass, which often leads to CoT derailment, i.e., the reasoning trajectory drifting off course due to compounding errors. This problem is particularly severe for smaller LMs with long CoTs due to their limited capacity. To address this, we analyze raw long CoTs and uncover a reasoning hierarchy consisting of planning and execution steps. Our analysis reveals that most reasoning errors stem from incorrect planning. Motivated by this observation, we propose Multi-Path Plan Aggregation (MPPA), a framework that augments single-pass reasoning with plan exploration and aggregation. Following a variable interval schedule based on the token position, MPPA generates multiple candidate plans and aggregates them into a refined planning step. To maintain efficiency, we adopt a minimal design in which the base LM serves as the primary policy, while a lightweight LoRA module implements the plan aggregation policy. We further observe that outcome-reward RL is inefficient for long trajectories (e.g., exceeding 4K tokens). To overcome this, we introduce online Step-DPO, a process-level preference optimization scheme that leverages Twisted Sequential Monte Carlo (TSMC) to provide scalable stepwise supervision using small LMs. This yields more efficient training, improved stability, and higher accuracy. Extensive experiments on challenging math, science, and logical reasoning benchmarks demonstrate that, with only 10% SFT data and 5% of preference pairs, our method outperforms both the DeepSeek-R1 distillation baseline and the outcome-reward RL baseline across multiple base models and tasks.

Related papers

• Why Reasoning Fails to Plan: A Planning-Centric Analysis of Long-Horizon Decision Making in LLM Agents [42.09897801169138]
  Large language model (LLM)-based agents exhibit strong step-by-step reasoning capabilities over short horizons, yet often fail to sustain coherent behavior over long planning horizons.<n>We argue that step-wise reasoning induces a form of step-wise greedy policy that is adequate for short horizons but fails in long-horizon planning.<n>We introduce FLARE as a minimal instantiation of future-aware planning to enforce explicit lookahead, value propagation, and limited commitment in a single model.
  arXiv  Detail & Related papers  (2026-01-29T20:52:32Z)
• Save the Good Prefix: Precise Error Penalization via Process-Supervised RL to Enhance LLM Reasoning [59.76691952347156]
  Reinforcement learning (RL) has emerged as a powerful framework for improving the reasoning capabilities of large language models (LLMs)<n>Most existing RL approaches rely on sparse outcome rewards, which fail to credit correct intermediate steps in partially successful solutions.<n>We propose Verifiable Prefix Policy Optimization (VPPO), which uses PRMs only to localize the first error during RL.
  arXiv  Detail & Related papers  (2026-01-26T21:38:20Z)
• Plan Then Action:High-Level Planning Guidance Reinforcement Learning for LLM Reasoning [22.177866778776814]
  We propose a two-stage framework designed to improve both high-level planning and fine-grained Chain-of-Thought (CoT) reasoning.<n>In the first stage, we leverage advanced LLMs to distill CoT into compact high-level guidance, which is then used for supervised fine-tuning.<n>In the second stage, we introduce a guidance-aware RL method that jointly optimize the final output and the quality of high-level guidance.
  arXiv  Detail & Related papers  (2025-10-02T09:28:13Z)
• From Long to Short: LLMs Excel at Trimming Own Reasoning Chains [48.692414597960244]
  O1/R1 style large reasoning models (LRMs) signal a substantial leap forward over conventional instruction-following LLMs.<n>Recent studies show that LRMs are prone to suffer from overthinking.<n>We propose a test-time scaling method, EDIT, which efficiently guides LRMs to identify the shortest correct reasoning paths at test time.
  arXiv  Detail & Related papers  (2025-09-07T19:00:44Z)
• Stop Spinning Wheels: Mitigating LLM Overthinking via Mining Patterns for Early Reasoning Exit [114.83867400179354]
  Overthinking can degrade overall performance of large language models.<n>We categorize reasoning into three stages: insufficient exploration stage, compensatory reasoning stage, and reasoning convergence stage.<n>We develop a lightweight thresholding strategy based on rules to improve reasoning accuracy.
  arXiv  Detail & Related papers  (2025-08-25T03:17:17Z)
• Pruning Long Chain-of-Thought of Large Reasoning Models via Small-Scale Preference Optimization [26.462701299259248]
  Large Reasoning Models (LRMs) have demonstrated strong performance on complex tasks through long Chain-of-Thought (CoT) reasoning.<n>Their lengthy outputs increase computational costs and may lead to overthinking, raising challenges in balancing reasoning effectiveness and efficiency.<n>This paper investigates efficient methods to reduce the generation length of LRMs.
  arXiv  Detail & Related papers  (2025-08-13T20:00:09Z)
• PLAN-TUNING: Post-Training Language Models to Learn Step-by-Step Planning for Complex Problem Solving [66.42260489147617]
  We introduce PLAN-TUNING, a framework that distills synthetic task decompositions from large-scale language models.<n>Plan-TUNING fine-tunes smaller models via supervised and reinforcement-learning objectives to improve complex reasoning.<n>Our analysis demonstrates how planning trajectories improves complex reasoning capabilities.
  arXiv  Detail & Related papers  (2025-07-10T07:30:44Z)
• CRISP: Complex Reasoning with Interpretable Step-based Plans [15.656686375199921]
  We introduce CRISP (Complex Reasoning with Interpretable Step-based Plans), a dataset of high-level plans for mathematical reasoning and code generation.<n>We demonstrate that fine-tuning a small model on CRISP enables it to generate higher-quality plans than much larger models using few-shot prompting.
  arXiv  Detail & Related papers  (2025-07-09T11:40:24Z)
• SmartThinker: Learning to Compress and Preserve Reasoning by Step-Level Length Control [5.224609066309358]
  Large reasoning models (LRMs) have exhibited remarkable reasoning capabilities through inference-time scaling.<n>Previous work has attempted to mitigate this issue by penalizing the overall length of generated samples during reinforcement learning.<n>We propose SmartThinker, a two-stage learnable framework designed to enable fine-grained control over the length of reasoning chains.
  arXiv  Detail & Related papers  (2025-07-06T11:21:47Z)
• When More is Less: Understanding Chain-of-Thought Length in LLMs [51.631483479081645]
  Large Language Models (LLMs) employ Chain-of-Thought (CoT) reasoning to deconstruct complex problems.<n>This paper argues that longer CoTs are often presumed superior, arguing that longer is not always better.
  arXiv  Detail & Related papers  (2025-02-11T05:28:59Z)
• VinePPO: Refining Credit Assignment in RL Training of LLMs [66.80143024475635]
  We propose VinePPO, a straightforward approach that leverages the flexibility of language environments to compute unbiased Monte Carlo-based estimates.<n>Our method consistently outperforms PPO and other baselines across MATH and GSM8K datasets in less wall-clock time.
  arXiv  Detail & Related papers  (2024-10-02T15:49:30Z)
• Tree-Planner: Efficient Close-loop Task Planning with Large Language Models [63.06270302774049]
  Tree-Planner reframes task planning with Large Language Models into three distinct phases. Tree-Planner achieves state-of-the-art performance while maintaining high efficiency.
  arXiv  Detail & Related papers  (2023-10-12T17:59:50Z)
• Guiding Language Model Reasoning with Planning Tokens [122.43639723387516]
  Large language models (LLMs) have recently attracted considerable interest for their ability to perform complex reasoning tasks. We propose a hierarchical generation scheme to encourage a more structural generation of chain-of-thought steps. Our approach requires a negligible increase in trainable parameters (0.001%) and can be applied through either full fine-tuning or a more parameter-efficient scheme.
  arXiv  Detail & Related papers  (2023-10-09T13:29:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.