Pith. sign in

Accelerating inference in large language models with a unified layer skipping strategy,

Yijin Liu, Fandong Meng, Jie Zhou, “Accelerating inference in large language models with a unified layer skipping strategy,” arXiv preprint arXiv:2404 · 2024 · arXiv 2404.06954

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
read on arXiv browse 1 citing papers

fields

cs.CL 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

Experience-Driven Dynamic Exits for LLMs with Reinforcement Learning

cs.CL · 2026-06-02 · unverdicted · novelty 6.0

LEDE reframes speculative decoding as an MDP and applies offline RL to learn dynamic policies for exit layer and speculation length selection, delivering 2.0-2.7x speedups over autoregressive decoding on Llama-2/3 models.

citing papers explorer

Showing 1 of 1 citing paper.

  • Experience-Driven Dynamic Exits for LLMs with Reinforcement Learning cs.CL · 2026-06-02 · unverdicted · none · ref 22

    LEDE reframes speculative decoding as an MDP and applies offline RL to learn dynamic policies for exit layer and speculation length selection, delivering 2.0-2.7x speedups over autoregressive decoding on Llama-2/3 models.