Early Stopping

Follow this topic Ask about this topic

Latest papers 40

5 papers in the last four weeks, level with the four weeks before. 0.0% of all new papers.

All topics

TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping
Apr 22, 2026Yannis Belkhiter, Seshu Tirupathi, Giulio Zizzo +1LLM Inference Efficiency Early Stopping
IBM Research Europe · Trinity College Dublin · Dublin, Ireland +1
ILDR: Geometric Early Detection of Grokking
Apr 22, 2026Shreel GolwalaNeural Network Generalization Early Stopping
Virginia Tech
Random Matrix Theory of Early-Stopped Gradient Flow: A Transient BBP Scenario
Apr 20, 2026Florentin Coeurdoux, Grégoire Ferré, Jean-Philippe BouchaudRandom Matrix Theory Early Stopping
1Capital Fund Management, Paris, France · 2X-CFM Chair of EconophysiX, Ecole polytechnique, Palaiseau, France · 3Académie des Sciences, Paris, France
Towards E-Value Based Stopping Rules for Bayesian Deep Ensembles
Apr 20, 2026Emanuel Sommer, Rickmer Schulte, Sarah Deubner +2Markov Chain Monte Carlo Early Stopping
LMU Munich, MCML
Early Stopping for Large Reasoning Models via Confidence Dynamics
Apr 6, 2026Parsa Hosseini, Sumit Nawathe, Mahdi Salmani +2Early Stopping Efficient Language Model Reasoning
University of Maryland · University of Southern California
EMS: Multi-Agent Voting via Efficient Majority-then-Stopping
Apr 3, 2026Yiqing Liu, Hantao Yao, Wu Liu +1Multi-Agent Reasoning Early Stopping
University of Science and Technology of China
Don't stop me now: How Validation Criteria Affect Checkpoint Selection and Early Stopping
Feb 25, 2026Andrea Apicella, Francesco Isgrò, Andrea Pollastro +1Early Stopping Neural Network Training Dynamics
Department of Information Engineering, Electrical Engineering, and Applied Mathematics (DIEM), University of Salerno, Via Giovanni Paolo II, 132, Fisciano (Salerno), 84084, Italy. · Department of Electrical Engineering and Information Technology, University of Naples Federico II, Via Claudio 21, Naples, 80125, Italy.
Beyond Accuracy: A Unified Random Matrix Theory Diagnostic Framework for Crash Classification Models
Feb 23, 2026Ibne Farabi Shihab, Sanjeda Akter, Anuj SharmaModel Selection Classification
Department of Computer Science, Iowa State University, Ames, IA 50010 · Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50010
Basic Inequalities for First-Order Optimization with Applications to Statistical Risk Analysis
Dec 31, 2025Seunghoon Paik, Kangjie Zhou, Matus Telgarsky +1Early Stopping Gradient Descent
University of California, Berkeley · Columbia University · New York University
On Regularization via Early Stopping for Least Squares Regression
Jun 6, 2024Rishi Sonthalia, Jackie Lok, Elizaveta RebrovaEarly Stopping Linear Regression
Department of Mathematics, Boston College · ORFE Department, Princeton University

Apr 22, 2026J/K move · Enter open · S save

TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping

Yannis Belkhiter, Seshu Tirupathi, Giulio Zizzo, John D. Kelleher

IBM Research Europe · Trinity College Dublin · Dublin, Ireland+1

LLM Inference Efficiency Early Stopping Efficient Language Model Reasoning Efficient Language Model Inference LLM Reasoning Large Reasoning Models+1

Open paper AskPDF arXiv

The field of Language Reasoning Models (LRMs) has been very active over the past few years with advances in training and inference techniques enabling LRMs to reason longer, and more accurately. However, a growing body of studies show that LRMs are still inefficient, over-generating verification and reflection steps. Additionally, the high-level role of each reasoning step and how different step types contribute to the generation of correct answers, is largely underexplored. To address this challenge, we introduce TRACES (Tagging of the Reasoning steps enabling Adaptive Cost-Efficient early-Stopping), a lightweight framework that tags reasoning steps in real-time, and enable adaptive, cost-efficient early stopping of large-language-model inferences. Building on this framework we monitor reasoning behaviors during inferences, and we find that LRMs tend to shift their reasoning behavior after reaching a correct answer. We demonstrate that the monitoring of the specific type of steps can produce effective interpretable early stopping criteria. We evaluate the TRACES framework on three mathematical reasoning benchmarks, namely, MATH500, GSM8K, AIME and two knowledge and reasoning benchmarks, MMLU and GPQA respectively. We achieve 20 to 50% token reduction while maintaining comparable accuracy to standard generation.