Weiran Huang

IDER: IDempotent Experience Replay for Reliable Continual Learning

Enhanced Continual Learning of Vision-Language Models with Model Fusion

Perceive Fast, Think Slow: A Cognitive-Inspired Framework for Time Series Analysis

Revisiting Visual Understanding in Multimodal Reasoning through a Lens of Image Perturbation

High-Power Training Data Identification with Provable Statistical Guarantees

RADAR: A Risk-Aware Dynamic Multi-Agent Framework for LLM Safety Evaluation via Role-Specialized Collaboration

Prototype Decoding and Re-Encoding for Class-Incremental Learning

Loss-Aligned Structured Pruning for Large Language Models

EFRame: Deeper Reasoning via Exploration-Filter-Replay Reinforcement Learning Framework

First SFT, Second RL, Third UPT: Continual Improving Multi-Modal LLM Reasoning via Unsupervised Post-Training

PANTHER: Generative Pretraining Beyond Language for Sequential User Behavior Modeling

Generalized Category Discovery via Reciprocal Learning and Class-Wise Distribution Regularization

Generalized Category Discovery Utilizing Reciprocal Learning and Class-wise Distribution Regularization

Understanding Grokking: Insights from Neural Network Robustness

Understanding Neural Tangent Kernel Dynamics Through Its Trace Evolution

Can I Understand What I Create? Self-Knowledge Evaluation of Large Language Models