← Search

Jiaojiao Lin

2 accepted papers

2026

PADD: Path-Aligned Decompression Distillation for Non-Router Teacher to Guide MoE Student Learning

ICML 2026poster

As large language models (LLMs) continue to scale, it becomes increasingly challenging to grow model capacity under fixed computation budgets. We propose Path-Aligned Decompression Distillation (PADD), a framework for distilling knowledge from dense teachers without explicit routing into mixture-of-…

Cited by 0SourceScholar