Latest Articles for Generalization

Poseidon uses machine learning to efficiently predict solutions for complex PDEs.

A new method for fine-tuning language models using self-attention.

Research shows diverse instructions improve language model performance in unseen tasks.

This study proposes a new approach to maintain learning in AI systems.

Introducing an innovative model to improve cardinality estimation accuracy in database queries.

This article examines how planning budgets affect DNC models in solving problems.

Introducing a universal framework for sharpness measures in machine learning.

Exploring how agents can adapt learned skills to new tasks using policy generators.

Explore how words change meaning over time and the factors influencing this process.

New techniques improve generalization in interconnected data models across various domains.

This article discusses how models can forget biases to improve predictions.

Examining how random feature models and Transformers handle unseen data.

This article discusses how languages change through iterative learning methods.

This article explores strategies for improving model generalization and understanding gradient behavior.

This article examines challenges in link prediction due to data distribution shifts.

A look into how LLMs tackle programming by example challenges.

New GNN architecture improves predictions through enhanced expressiveness and fragment integration.

A new method improves model adaptability across domains using prompt learning and gradient alignment.

A new method uses natural language explanations to improve entity matching.

FouRA enhances image generation by improving quality and diversity.

Learn how bagging boosts model performance across various applications.

A study highlights gaps in reasoning abilities of LLMs for math problem solving.

PartCLIPSeg improves part segmentation in computer vision for better object recognition.

Study analyzes generalization and performance of random feature ridge regression using eigenvalues.

Exploring the role of transformers in predicting sequential data outcomes.

SiT enhances agents' ability to generalize in reinforcement learning through symmetry and attention.

New methods improve speech models for languages with limited data.

A look at the role of complexity in model performance.

A novel loss function enhances feature learning in classification tasks.

New kernels enhance data analysis through complex function interactions.

New neural operators improve accuracy in predicting magnetic material behavior.

Examining vulnerabilities from clean-label backdoor attacks and how generalization bounds can help.

Weight clipping enhances model performance in deep learning and reinforcement learning.

SLIMER enhances NER performance by focusing on definitions and guidelines.

NGOs utilize neural networks to simplify solving complex partial differential equations efficiently.

This research examines how language models predict the order of adjectives.

GFlowNets enable effective sampling from complex distributions with strong generalization capabilities.

A new method to improve model performance on out-of-distribution data.

An overview of methods to safeguard personal information while maintaining data usefulness.

A study on the learning capabilities of large language models in modular arithmetic tasks.