regularization
Regularization is a technique used to prevent overfitting in machine learning models by adding a penalty to the loss function that discourages overly complex models, thereby promoting generalization to unseen data.
- Adversarial Robustness of Nonparametric Regression
- Efficient Algorithms for Robust and Partial Semi-Discrete Optimal Transport
- Energy Matching: Unifying Flow Matching and Energy-Based Models for Generative Modeling
- Evolving and Regularizing Meta-Environment Learner for Fine-Grained Few-Shot Class-Incremental Learning
- Flatness is Necessary, Neural Collapse is Not: Rethinking Generalization via Grokking
- Hankel Singular Value Regularization for Highly Compressible State Space Models
- Learning to Instruct for Visual Instruction Tuning
- Less is More: Unlocking Specialization of Time Series Foundation Models via Structured Pruning
- Localist Topographic Expert Routing: A Barrel Cortex-Inspired Modular Network for Sensorimotor Processing
- Neural Collapse in Cumulative Link Models for Ordinal Regression: An Analysis with Unconstrained Feature Model
- Normalize Filters! Classical Wisdom for Deep Vision
- On Extending Direct Preference Optimization to Accommodate Ties
- PROFIT: A Specialized Optimizer for Deep Fine Tuning
- Statistical Inference for Gradient Boosting Regression
- Steering Information Utility in Key-Value Memory for Language Model Post-Training
- Temperature is All You Need for Generalization in Langevin Dynamics and other Markov Processes
- Temporal-Difference Variational Continual Learning