2 papers
cs.LG2026
Share First, Route What Remains: A Unified Framework for Token-Adaptive MoE Computation
Gongli Zhang, Zhulin Liu, C. L. Philip Chen
Mixture-of-experts (MoE) models have recently moved beyond routing a fixed number of complete experts. Shared-expert designs preserve reusable knowledge, fine-grained methods vary…
cs.LG2024
Incremental Self-training for Semi-supervised Learning
Jifeng Guo, Zhulin Liu, Tong Zhang +1
Semi-supervised learning provides a solution to reduce the dependency of machine learning on labeled data. As one of the efficient semi-supervised techniques, self-training (ST) ha…