Latest Articles for Machine Learning

Fed-RAA enhances federated learning by adapting to client resources for faster training.

Examining the impact of data contamination on language model performance and evaluation.

A new method improves example selection and instruction optimization for large language models.

Aurora improves query rewriting using Equality Saturation and Reinforcement Learning.

A look at controlling language model behavior with the KL-then-steer technique.

PE-Rank improves passage ranking efficiency with single passage embeddings.

A new approach in federated learning captures data dependencies while ensuring privacy.

Strategies to manage performance issues during continual pre-training of large language models.

Researchers combine data techniques to model complex systems effectively.

A new benchmark for machine unlearning enhances evaluation and comparison of methods.

How fine-tuning affects language models' ability to recall facts accurately.

A new method enhances accuracy in calculating sublimation enthalpies for molecular crystals.

A new method improves image generation without paired datasets.

Discover how enhanced machine learning helps improve composite material production.

A new method combines unsupervised learning with combinatorial optimization to improve decision-making.

SNAPS alerts scientists to unusual asteroid events for further study.

This article explores how adversaries impact teamwork among language models.

A new metric improves evaluation of text classification models across different domains.

A detailed look at Sibson's -mutual information and its multifaceted applications.

New methods enhance optimization in complex scenarios using latent spaces and noise.

Data contamination affects the evaluation of large language models significantly.

A new model enhances node classification by leveraging related knowledge.

A new method for enhancing machine learning through suboptimal examples and human feedback.

This article examines how LLMs can generate synthetic tables effectively while addressing key challenges.

Learn how Bayesian inference enhances neural networks and decision-making.

Mamba’s context-extension method improves handling of long sequences without additional training.

Smaller models can learn effectively from larger models' reasoning steps.

This article examines methods to remove user data from neural networks efficiently.

RFM improves data generation by managing boundary constraints effectively.

Highlighting the importance of data in training large language models.

A system that optimizes machine learning predictions through model cascades.

QuEE combines quantization and early exiting for efficient machine learning.

Exploring the role of transformers in predicting sequential data outcomes.

New methods improve accuracy in recognizing overlapping sounds across diverse audio sources.

A novel approach to fair data selection in real-time streams.

A study on how pose estimation models perform with corrupted images.

Ensuring AI explanations are clear and reliable for user trust.

A new approach improves optimization of complex loss functions in neural networks.

Introducing iWISDM, a new benchmark for assessing machine instruction following with visuals and language.

A new benchmark enhances long-term predictions in event sequences across industries.