Cover Image

Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Min, Dehai, Vaccarino, Giovanni, Chen, Huiyi, Wu, Yongliang, Yona, Gal, Cheng, Lu
Format:	Preprint
Published:	2026
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Stop When Enough: Adaptive Early-Stopping for Chain-of-Thought Reasoning
by: Sun, Renliang, et al.
Published: (2025)

Answer Convergence as a Signal for Early Stopping in Reasoning
by: Liu, Xin, et al.
Published: (2025)

Latent Thoughts Tuning: Bridging Context and Reasoning with Fused Information in Latent Tokens
by: Liu, Weihao, et al.
Published: (2026)

When Is Thinking Enough? Early Exit via Sufficiency Assessment for Efficient Reasoning
by: Xiang, Yang, et al.
Published: (2026)

Dynamic Early Exit in Reasoning Models
by: Yang, Chenxu, et al.
Published: (2025)

TERMINATOR: Learning Optimal Exit Points for Early Stopping in Chain-of-Thought Reasoning
by: Nagle, Alliot, et al.
Published: (2026)

NEAT: Neuron-Based Early Exit for Large Reasoning Models
by: Liu, Kang, et al.
Published: (2026)

Keep Guessing? When Considering Inference Scaling, Mind the Baselines
by: Yona, Gal, et al.
Published: (2024)

FlashThink: An Early Exit Method For Efficient Reasoning
by: Jiang, Guochao, et al.
Published: (2025)

Can Large Language Models Faithfully Express Their Intrinsic Uncertainty in Words?
by: Yona, Gal, et al.
Published: (2024)

Adaptive Stopping for Multi-Turn LLM Reasoning
by: Zhou, Xiaofan, et al.
Published: (2026)

SpecExit: Accelerating Large Reasoning Model via Speculative Exit
by: Yang, Rubing, et al.
Published: (2025)

When to Memorize and When to Stop: Gated Recurrent Memory for Long-Context Reasoning
by: Sheng, Leheng, et al.
Published: (2026)

Early Stopping for Large Reasoning Models via Confidence Dynamics
by: Hosseini, Parsa, et al.
Published: (2026)

TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping
by: Belkhiter, Yannis, et al.
Published: (2026)

The Zero-Step Thinking: An Empirical Study of Mode Selection as Harder Early Exit in Reasoning Models
by: Tan, Yuqiao, et al.
Published: (2025)

Hallucinations Undermine Trust; Metacognition is a Way Forward
by: Yona, Gal, et al.
Published: (2026)

Narrowing the Knowledge Evaluation Gap: Open-Domain Question Answering with Multi-Granularity Answers
by: Yona, Gal, et al.
Published: (2024)

QuCo-RAG: Quantifying Uncertainty from the Pre-training Corpus for Dynamic Retrieval-Augmented Generation
by: Min, Dehai, et al.
Published: (2025)

Semantic Deception: When Reasoning Models Can't Compute an Addition
by: de Leeuw, Nathaniël, et al.
Published: (2025)

EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering and Reasoning
by: Wei, Mingyang, et al.
Published: (2026)

SimpleQA Verified: A Reliable Factuality Benchmark to Measure Parametric Knowledge
by: Haas, Lukas, et al.
Published: (2025)

Thinking Out of Order: When Output Order Stops Reflecting Reasoning Order in Diffusion Language Models
by: Yu, Longxuan, et al.
Published: (2026)

Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
by: Sui, Yang, et al.
Published: (2025)

Language as a Latent Variable for Reasoning Optimization
by: Wu, Linjuan, et al.
Published: (2026)

Active Confusion Expression in Large Language Models: Leveraging World Models toward Better Social Reasoning
by: Du, Jialu, et al.
Published: (2025)

Early Exit Is a Natural Capability in Transformer-based Models: An Empirical Study on Early Exit without Joint Optimization
by: Shan, Weiqiao, et al.
Published: (2024)

When to Reason: Semantic Router for vLLM
by: Wang, Chen, et al.
Published: (2025)

Know When To Stop: A Study of Semantic Drift in Text Generation
by: Spataru, Ava, et al.
Published: (2024)

Early-Exit and Instant Confidence Translation Quality Estimation
by: Zouhar, Vilém, et al.
Published: (2025)

CoT Vectors: Transferring and Probing the Reasoning Mechanisms of LLMs
by: Li, Li, et al.
Published: (2025)

Stop Rewarding Hallucinated Steps: Faithfulness-Aware Step-Level Reinforcement Learning for Small Reasoning Models
by: Nie, Shuo, et al.
Published: (2026)

Pause or Fabricate? Training Language Models for Grounded Reasoning
by: Qiu, Yiwen, et al.
Published: (2026)

Evaluating Social Bias in RAG Systems: When External Context Helps and Reasoning Hurts
by: Parihar, Shweta, et al.
Published: (2026)

Your Models Have Thought Enough: Training Large Reasoning Models to Stop Overthinking
by: Han, Jinyi, et al.
Published: (2025)

LYNX: Learning Dynamic Exits for Confidence-Controlled Reasoning
by: Akgül, Ömer Faruk, et al.
Published: (2025)

STaR-SQL: Self-Taught Reasoner for Text-to-SQL
by: He, Mingqian, et al.
Published: (2025)

Stop Before You Fail: Operational Capability Boundaries for Mitigating Unproductive Reasoning in Large Reasoning Models
by: Zhang, Qingjie, et al.
Published: (2025)

ADEPT: Adaptive Dynamic Early-Exit Process for Transformers
by: Yoo, Sangmin, et al.
Published: (2026)

Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge Reasoning
by: Chen, Yongrui, et al.
Published: (2025)