Hai Huang

LLM-JEPA: Large Language Models Meet Joint Embedding Predictive Architectures

Directed Information $\gamma$-covering: An Information-Theoretic Framework for Context Engineering

Next Token Perception Score: Analytical Assessment of your LLM Perception Skills

LoRA Users Beware: A Few Spurious Tokens Can Manipulate Your Finetuned Model

Guided Score identity Distillation for Data-Free One-Step Text-to-Image Generation

Adversarial Score identity Distillation: Rapidly Surpassing the Teacher in One Step

Relative Preference Optimization: Enhancing LLM Alignment through Contrasting Responses across Identical and Diverse Prompts

ALLoRA: Adaptive Learning Rate Mitigates LoRA Fatal Flaws