Efficiently Aligning Draft Models via Parameter- and Data-Efficient Adaptation

EDA adapts draft models to fine-tuned LLMs via lightweight private components, regenerated training data, and selective sampling, restoring speculative decoding performance at much lower cost than full retraining.

Luxi Lin, Zhihang Lin, Zhanpeng Zeng, Yuhao Chen and 3 more

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026