Results 281 to 290 of about 12,738,743 (318)

Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoning

International Conference on Learning Representations
Self-consistency (SC) has been a widely used decoding strategy for chain-of-thought reasoning. Despite bringing significant performance improvements across a variety of multi-step reasoning tasks, it is a high-cost method that requires multiple sampling ...
Yi-Wei Li, Pei-Wen Yuan, Shaoxiong Feng, Bo-Yuan Pan, Xing-Lin Wang, Bin Sun, Heda Wang, Kan Li   +7 more
semanticscholar   +1 more source

Benefits of Early Stopping in Gradient Descent for Overparameterized Logistic Regression

International Conference on Machine Learning
In overparameterized logistic regression, gradient descent (GD) iterates diverge in norm while converging in direction to the maximum $\ell_2$-margin solution -- a phenomenon known as the implicit bias of GD.
Jingfeng Wu, Peter L. Bartlett, Matus Telgarsky, Bin Yu   +3 more
semanticscholar   +1 more source

Stopping a trial early – and then what?

Clinical Trials, 2012
Background This article addresses a problem arising when a trial shows such strong evidence of benefit of the tested intervention that it stops early with an observed effect size for the experimental treatment that is statistically significantly better than the control. Within the classical frequentist framework
openaire   +2 more sources

On Regularization via Early Stopping for Least Squares Regression

arXiv.org
A fundamental problem in machine learning is understanding the effect of early stopping on the parameters obtained and the generalization capabilities of the model.
Rishi Sonthalia, Jackie Lok, E. Rebrova
semanticscholar   +1 more source

Don't Waste Your Time: Early Stopping Cross-Validation

AutoML
State-of-the-art automated machine learning systems for tabular data often employ cross-validation; ensuring that measured performances generalize to unseen data, or that subsequent ensembling does not overfit.
E. Bergman, L. Purucker, Frank Hutter
semanticscholar   +1 more source

Modelling and simulating early stopping of RCTs: a case study of early stop due to harm

Journal of Experimental & Theoretical Artificial Intelligence, 2012
Despite efforts from regulatory agencies (e.g. NIH, FDA), recent systematic reviews of randomised controlled trials (RCTs) show that top medical journals continue to publish trials without requiring authors to report details for readers to evaluate early stopping decisions carefully.
openaire   +2 more sources

The Dangers of Stopping a Trial Too Early

Journal of Bone and Joint Surgery, 2012
Abstract: To ensure that participants in randomized controlled trials are protected from harm, interim analyses and review of results by an independent data monitoring committee have become standard practice. If an analysis of accumulating data partway through a trial reveals an unanticipated degree of benefit or toxicity, or ...
Matthias, Briel, Dirk, Bassler, Amy T, Wang, Gordon H, Guyatt, Victor M, Montori   +4 more
openaire   +2 more sources

Reflections on early stopping of a clinical trial

American Heart Journal, 2006
Large mortality-based trials in cardiovascular disease have made a major impact on the care of afflicted patients over the past 2 decades. Several developments have conspired to make their conduct and completion increasingly complicated. These include greater emphasis on efficient clinical care, the advent of multiple evidence-based therapies as ...
Paul W, Armstrong, Christopher B, Granger   +1 more
openaire   +2 more sources

Interpretation of Clinical Trials That Stopped Early

JAMA, 2016
Clinical trials require significant resources to complete in terms of patients, investigators, and time and should be carefully designed and conducted so that they use the minimum amount of resources necessary to answer the motivating clinical question.
Kert, Viele, Anna, McGlothlin, Kristine, Broglio   +2 more
openaire   +2 more sources

AdaEDL: Early Draft Stopping for Speculative Decoding of Large Language Models via an Entropy-based Lower Bound on Token Acceptance Probability

ENLSP
Speculative decoding is a powerful technique that attempts to circumvent the autoregressive constraint of modern Large Language Models (LLMs). The aim of speculative decoding techniques is to improve the average inference time of a large, target model ...
Sudhanshu Agrawal, Wonseok Jeon, Min-Gu Lee   +2 more
semanticscholar   +1 more source