pith. sign in

arxiv: 2502.06772 · v2 · pith:23GDE64Gnew · submitted 2025-02-10 · 💻 cs.CL · cs.AI· cs.LG

ReasonFlux: Hierarchical LLM Reasoning via Scaling Thought Templates

classification 💻 cs.CL cs.AIcs.LG
keywords reasoningthoughttemplateshierarchicalscalingmatho1-previewproblems
0
0 comments X
read the original abstract

We present that hierarchical LLM reasoning via scaling thought templates can effectively optimize the reasoning search space and outperform the mathematical reasoning capabilities of powerful LLMs like OpenAI o1-preview and DeepSeek V3. We train our ReasonFlux-32B model with only 8 GPUs and introduces three innovations: (i) a structured and generic thought template library, containing around 500 high-level thought templates capable of generalizing to similar or relevant reasoning problems; (ii) performing hierarchical reinforcement learning on a sequence of thought templates instead of long CoTs, optimizing a base LLM to plan out an optimal template trajectory for gradually handling complex problems; (iii) a brand new inference scaling system that enables hierarchical LLM reasoning by adaptively scaling thought templates at inference time. With a template trajectory containing more explainable reasoning structures than DeepSeek-R1 and o3-mini, our ReasonFlux-32B significantly advances math reasoning capabilities to state-of-the-art levels. Notably, on the MATH benchmark, it achieves an accuracy of 91.2% and surpasses o1-preview by 6.7%. On the USA Math Olympiad (AIME) benchmark, ReasonFlux-32B solves an average of 56.7% of problems, surpassing o1-preview and DeepSeek-V3 by 27% and 45%, respectively. Code: https://github.com/Gen-Verse/ReasonFlux

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Capacity, Not Format: Rethinking Structured Reasoning Failures

    cs.AI 2026-06 unverdicted novelty 7.0

    Empirical study across 4 models and 5 benchmarks finds that structured output formats degrade LLM reasoning performance primarily in capacity-limited models, with recovery via delayed formatting.

  2. POSTCONDBENCH: Benchmarking Correctness and Completeness in Formal Postcondition Inference

    cs.SE 2026-05 unverdicted novelty 7.0

    POSTCONDBENCH is a new multilingual benchmark that evaluates LLM postcondition generation on real code using defect discrimination to assess completeness beyond surface matching.

  3. Validity-Calibrated Reasoning Distillation

    cs.LG 2026-04 unverdicted novelty 7.0

    Validity-calibrated reasoning distillation improves transfer of reasoning skills by modulating updates based on relative local validity of next steps instead of enforcing full trajectory imitation.

  4. Validity-Calibrated Reasoning Distillation

    cs.LG 2026-04 unverdicted novelty 7.0

    Validity-calibrated reasoning distillation improves small LLMs by using relative local validity of next steps to dynamically adjust imitation strength instead of enforcing full trajectory matching.

  5. Procedural Knowledge at Scale Improves Reasoning

    cs.CL 2026-04 unverdicted novelty 6.0

    Reasoning Memory decomposes reasoning trajectories into 32 million subquestion-subroutine pairs and retrieves them via in-thought prompts to improve language model performance on math, science, and coding benchmarks b...

  6. The Landscape of Agentic Reinforcement Learning for LLMs: A Survey

    cs.AI 2025-09 accept novelty 6.0

    Survey that defines agentic RL for LLMs via POMDPs, introduces a taxonomy of planning/tool-use/memory/reasoning capabilities and domains, and compiles open environments from over 500 papers.

  7. Learning to Refine: Self-Refinement of Parallel Reasoning in LLMs

    cs.LG 2025-08 conditional novelty 6.0

    GSR jointly trains LLMs to generate candidate solutions and refine a superior final answer from them, achieving state-of-the-art performance on five mathematical benchmarks while transferring across model scales.

  8. MMaDA: Multimodal Large Diffusion Language Models

    cs.CV 2025-05 unverdicted novelty 6.0

    MMaDA is a unified multimodal diffusion model using mixed chain-of-thought fine-tuning and a new UniGRPO reinforcement learning algorithm that outperforms specialized models in reasoning, understanding, and text-to-im...

  9. The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes

    cs.CL 2026-06 unverdicted novelty 4.0

    A literature survey that introduces a taxonomy for LLM reasoning paradigms, analyzes methodological trends, and synthesizes failure modes from over 300 papers.

  10. From System 1 to System 2: A Survey of Reasoning Large Language Models

    cs.AI 2025-02 accept novelty 3.0

    The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.