Entropy After $\langle \texttt{/Think} \rangle$ for reasoning model early exiting

• URL: http://arxiv.org/abs/2509.26522v1
• Date: Tue, 30 Sep 2025 16:59:37 GMT
• Title: Entropy After $\langle \texttt{/Think} \rangle$ for reasoning model early exiting
• Authors: Xi Wang, James McInerney, Lequn Wang, Nathan Kallus,
• Abstract summary: We show that large reasoning models overthink, continuing to revise answers even after reaching the correct solution.<n>We propose Entropy After /Think> (EAT) for monitoring and deciding whether to exit reasoning early.<n>EAT reduces token usage by 13 - 21% without harming accuracy.
• Score: 38.93424884988798
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large reasoning models show improved performance with longer chains of thought. However, recent work has highlighted (qualitatively) their tendency to overthink, continuing to revise answers even after reaching the correct solution. We quantitatively confirm this inefficiency by tracking Pass@1 for answers averaged over a large number of rollouts and find that the model often begins to always produce the correct answer early in the reasoning, making extra reasoning a waste of tokens. To detect and prevent overthinking, we propose a simple and inexpensive novel signal -- Entropy After </Think> (EAT) -- for monitoring and deciding whether to exit reasoning early. By appending a stop thinking token (</think>) and monitoring the entropy of the following token as the model reasons, we obtain a trajectory that decreases and stabilizes when Pass@1 plateaus; thresholding its variance under an exponential moving average yields a practical stopping rule. Importantly, our approach enables adaptively allocating compute based on the EAT trajectory, allowing us to spend compute in a more efficient way compared with fixing the token budget for all questions. Empirically, on MATH500 and AIME2025, EAT reduces token usage by 13 - 21% without harming accuracy, and it remains effective in black box settings where logits from the reasoning model are not accessible, and EAT is computed with proxy models.

Related papers

• ENTRA: Entropy-Based Redundancy Avoidance in Large Language Model Reasoning [30.786062954495403]
  Large Reasoning Models (LRMs) often suffer from overthinking, generating unnecessarily long reasoning chains even for simple tasks.<n>We propose ENTRA, an entropy-based training framework that suppresses redundant reasoning while preserving performance.
  arXiv  Detail & Related papers  (2026-01-12T01:26:30Z)
• Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production [55.76222360698305]
  We explore a class of supervised training objectives that allow a language model to dynamically and autonomously scale the number of compute steps used for each input token.<n>For any token, the model can request additional compute steps by emitting a don't know> output.<n>We find that the CYB model requests additional steps when doing so improves accuracy, and the model adapts its processing time to token-level complexity and context.
  arXiv  Detail & Related papers  (2025-10-13T21:07:05Z)
• One Token Embedding Is Enough to Deadlock Your Large Reasoning Model [91.48868589442837]
  We present the Deadlock Attack, a resource exhaustion method that hijacks an LRM's generative control flow.<n>Our method achieves a 100% attack success rate across four advanced LRMs.
  arXiv  Detail & Related papers  (2025-10-12T07:42:57Z)
• Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulation [82.62935304152239]
  Large Language Models (LLMs) have demonstrated remarkable reasoning abilities on complex problems using long Chain-of-Thought (CoT) reasoning.<n>They often suffer from overthinking, meaning generating unnecessarily lengthy reasoning steps for simpler problems.<n>We introduce a novel metric Token Entropy Cumulative Average (TECA), which measures the extent of exploration throughout the reasoning process.
  arXiv  Detail & Related papers  (2025-10-02T17:36:50Z)
• Entropy-Guided Loop: Achieving Reasoning through Uncertainty-Aware Generation [0.0]
  entropy-guided refinement is a lightweight, test-time loop that uses token-level uncertainty to trigger a single, targeted refinement pass.<n>We demonstrate that this uncertainty-aware loop provides an effective middle ground between single-pass inference and expensive reasoning chains.
  arXiv  Detail & Related papers  (2025-08-26T22:29:12Z)
• 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)
• Do Thinking Tokens Help or Trap? Towards More Efficient Large Reasoning Model [7.8354921036790275]
  Large Reasoning Models (LRMs) excel at solving complex problems but face an overthinking dilemma.<n>When handling simple tasks, they often produce verbose responses overloaded with thinking tokens.<n>These tokens trigger unnecessary high-level reasoning behaviors like reflection and backtracking, reducing efficiency.
  arXiv  Detail & Related papers  (2025-06-30T13:30:33Z)
• Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiency [24.56015832583054]
  Explicit self-reflection, signaled by tokens such as "Wait" and "Hmm", is necessary for advanced reasoning.<n>We propose NoWait, a simple yet effective approach that disables explicit self-reflection by suppressing these tokens during inference.
  arXiv  Detail & Related papers  (2025-06-10T01:54:04Z)
• CoThink: Token-Efficient Reasoning via Instruct Models Guiding Reasoning Models [56.40065909544213]
  Large language models (LLMs) benefit from increased test-time compute, a phenomenon known as test-time scaling.<n>However, reasoning-optimized models often overthink even simple problems, producing excessively verbose outputs and leading to low token efficiency.<n>We identify two key causes of this verbosity: (1) reinforcement learning reduces the information density of forward reasoning, and (2) backward chain-of thought training encourages redundant and often unnecessary verification steps.
  arXiv  Detail & Related papers  (2025-05-28T06:24:45Z)
• VeriThinker: Learning to Verify Makes Reasoning Model Efficient [52.74493506816969]
  Large Reasoning Models excel at complex tasks using Chain-of-Thought (CoT) reasoning.<n>Their tendency to overthinking leads to unnecessarily lengthy reasoning chains.<n>We introduce VeriThinker, a novel approach for CoT compression.
  arXiv  Detail & Related papers  (2025-05-23T14:17:56Z)
• DAST: Difficulty-Adaptive Slow-Thinking for Large Reasoning Models [30.184895117009457]
  This paper introduces Difficulty-Adaptive Slow Thinking (DAST), a novel framework that enables models to autonomously adjust the length of Chain-of-Thought (CoT) based on problem difficulty.<n>Experiments on diverse datasets and model scales demonstrate that DAST effectively mitigates overthinking while preserving reasoning accuracy on complex problems.
  arXiv  Detail & Related papers  (2025-03-06T14:23:06Z)
• O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning [98.3430004984531]
  We propose Length-Harmonizing Fine-Tuning (O1-Pruner) to minimize reasoning overhead while maintaining accuracy.<n>Our code is coming soon at https://github.com/StarDewXXX/O1-Pruner.
  arXiv  Detail & Related papers  (2025-01-22T01:35:11Z)

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.