Showing 2026Show all
3 papers · 1 filter
cs.LG2026
Why Ghost Outputs Teach: A Kernel-Based Understanding of Subliminal Learning
Zhe Li, Bicheng Ying, Chaosheng Dong +1
Subliminal Learning (SL) is a recently identified phenomenon in which a student model acquires downstream task capabilities by matching seemingly unrelated auxiliary outputs from a…
cs.LG2026
Learning Dynamics of Zeroth-Order Optimization: A Kernel Perspective
Zhe Li, Bicheng Ying, Zidong Liu +1
Classical optimization theory establishes that zeroth-order (ZO) algorithms suffer from a dimension-dependent slowdown, with convergence rates typically scaling with the model dime…
cs.LG2026
HOSL: Hybrid-Order Split Learning for Memory-Constrained Edge Training
Aakriti Lnu, Zhe Li, Dandan Liang +3
Split learning (SL) enables collaborative training of large language models (LLMs) between resource-constrained edge devices and compute-rich servers by partitioning model computat…