Relaxed Contrastive Learning for Federated Learning

HarmAug: Effective Data Augmentation for Knowledge Distillation of Safety Guard Models

Learning Diverse Attacks on Large Language Models for Robust Red-Teaming and Safety Tuning

Parameter Expanded Stochastic Gradient Markov Chain Monte Carlo

Lipsum-FT: Robust fine-Tuning of Zero-Shot Models Using Random Text Guidance

Stochastic Optimal Control for Diffusion Bridges in Function Spaces

Model Fusion through Bayesian Optimization in Language Model Fine-Tuning

Safeguard Text-to-Image Diffusion Models with Human Feedback Inversion