Step-GRPO: Internalizing Dynamic Early Exit for Efficient Reasoning

• URL: http://arxiv.org/abs/2604.16890v1
• Date: Sat, 18 Apr 2026 07:48:53 GMT
• Title: Step-GRPO: Internalizing Dynamic Early Exit for Efficient Reasoning
• Abstract summary: Step-GRPO internalizes dynamic early-exit capabilities directly into the model.<n>Our method reduces token consumption by 32.0% compared to the vanilla model.
• Score: 39.067556802856444
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large reasoning models that use long chain-of-thought excel at problem-solving yet waste compute on redundant checks. Curbing this overthinking is hard: training-time length penalties can cripple ability, while inference-time early-exit adds system overhead. To bridge this gap, we propose Step-GRPO, a novel post-training framework that internalizes dynamic early-exit capabilities directly into the model. Step-GRPO shifts the optimization objective from raw tokens to semantic steps by utilizing linguistic markers to structure reasoning. We introduce a Dynamic Truncated Rollout mechanism that exposes the model to concise high-confidence trajectories during exploration, synergized with a Step-Aware Relative Reward that dynamically penalizes redundancy based on group-level baselines. Extensive experiments across three model sizes on diverse benchmarks demonstrate that Step-GRPO achieves a superior accuracy-efficiency trade-off. On Qwen3-8B, our method reduces token consumption by 32.0\% compared to the vanilla model while avoiding the accuracy degradation observed in traditional length-penalty methods.

Related papers

• Stop When Further Reasoning Won't Help: Attention-State Adaptive Generation in Reasoning Models [11.158010513386666]
  Large reasoning models (LRMs) can solve complex problems through explicit chain-of-thought reasoning processes.<n>LRMs often suffer from overthinking, resulting in redundant token outputs and degraded accuracy.<n>We propose ASAG, which infers the model's reasoning state and adaptively adjusts the generation strategy.
  arXiv  Detail & Related papers  (2026-06-13T02:58:29Z)
• Stabilizing Efficient Reasoning with Step-Level Advantage Selection [54.563811052329235]
  We show that short-context post-training alone, using standard GRPO without any length-aware objective, already induces substantial reasoning compression.<n>We propose Step-level Advantage Selection (SAS), which operates at the reasoning-step level and assigns a zero advantage to low-confidence steps in correct rollouts and to high-confidence steps in verifier-failed rollouts.
  arXiv  Detail & Related papers  (2026-04-27T03:34:59Z)
• GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning [48.32628338889922]
  Group Relative Policy Optimization suffers from indiscriminate credit assignment for intermediate steps.<n>We introduce a model-free and verifiable process supervision via probing the model's belief in the correct answer.<n>This approach enables more targeted and sample-efficient policy updates.
  arXiv  Detail & Related papers  (2026-04-22T15:08:58Z)
• Stepwise Penalization for Length-Efficient Chain-of-Thought Reasoning [66.22060690012512]
  Large reasoning models improve with more test-time computation, but often overthink, producing unnecessarily long chains-of-thought that raise cost without improving accuracy.<n>We propose Step-wise Adaptive Penalization (SWAP), a fine-grained framework that allocates length reduction across steps based on intrinsic contribution.
  arXiv  Detail & Related papers  (2026-02-27T20:23:59Z)
• Difficulty-Estimated Policy Optimization [38.86673795561421]
  We propose Difficulty-Estimated Policy Optimization (DEPO), a novel framework designed to optimize the efficiency and robustness of reasoning alignment.<n>Our approach significantly lowers the computational barrier for training high-performance reasoning models, offering a more sustainable path for reasoning scaling.
  arXiv  Detail & Related papers  (2026-02-06T04:12:23Z)
• POP: Prefill-Only Pruning for Efficient Large Model Inference [5.743318651374061]
  Large Language Models (LLMs) and Vision-Language Models (VLMs) have demonstrated remarkable capabilities.<n>Existing structured pruning methods, while hardware-efficient, often suffer from significant accuracy degradation.<n>We argue that this failure stems from a stage-agnostic pruning approach that overlooks the asymmetric roles between the prefill and decode stages.
  arXiv  Detail & Related papers  (2026-02-03T09:22:26Z)
• PROMISE: Process Reward Models Unlock Test-Time Scaling Laws in Generative Recommendations [52.67948063133533]
  Generative Recommendation has emerged as a promising paradigm, reformulating recommendation as a sequence-to-sequence generation task over hierarchical Semantic IDs.<n>Existing methods suffer from a critical issue we term Semantic Drift, where errors in early, high-level tokens irreversibly divert the generation trajectory into irrelevant semantic subspaces.<n>We propose Promise, a novel framework that integrates dense, step-by-step verification into generative models.
  arXiv  Detail & Related papers  (2026-01-08T07:38:46Z)
• Addressing Overthinking in Large Vision-Language Models via Gated Perception-Reasoning Optimization [56.59356959631999]
  Gated Perception-Reasoning Optimization (GPRO) is a meta-reasoning controller that dynamically routes computation among three decision paths.<n>GPRO substantially improves both accuracy and efficiency, outperforming recent slow-thinking methods.
  arXiv  Detail & Related papers  (2026-01-07T23:05:17Z)
• Arbitrage: Efficient Reasoning via Advantage-Aware Speculation [71.45710345765528]
  Speculative Decoding accelerates inference by employing a fast but inaccurate draft model to autoregressively propose tokens.<n>But due to unnecessary rejections caused by token mismatches in semantically equivalent steps, traditional token-level Speculative Decoding struggles in reasoning tasks.<n>We propose Arbitrage, a novel step-level speculative generation framework that routes generation dynamically based on the relative advantage between draft and target models.
  arXiv  Detail & Related papers  (2025-12-04T17:50:53Z)
• Towards Flash Thinking via Decoupled Advantage Policy Optimization [11.025775055262569]
  Large Reasoning Models (LRMs) have achieved remarkable performance in solving complex problems via supervised fine-tuning (SFT) and reinforcement learning (RL)<n>Existing RL algorithms suffer from excessively lengthy responses and overthinking issues, resulting in increased inference latency and computational consumption.<n>We propose a novel RL framework, DEPO, to reduce inefficient reasoning for models.
  arXiv  Detail & Related papers  (2025-10-17T07:19:20Z)
• S-GRPO: Early Exit via Reinforcement Learning in Reasoning Models [2.9925837108958864]
  Test-Time Scaling emerges as an active research focus in the large language model community.<n>Recent studies reveal that reasoning models (even Qwen3) consistently exhibit excessive thought redundancy.<n>This paper introduces Serial-Group Decaying-Reward Policy Optimization (S-GRPO), a novel reinforcement learning paradigm.
  arXiv  Detail & Related papers  (2025-05-12T15:50:44Z)

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.