The Diminishing Returns of Early-Exit Decoding in Modern LLMs

Abstract

In Large Language Model (LLM) inference, early-exit refers to stopping computation at an intermediate layer once the prediction is sufficiently confident, thereby reducing latency and cost. However, recent LLMs adopt improved pretraining recipes and architectures that reduce layer redundancy, potentially limiting early-exit opportunities. We re-evaluate layer-wise early-exit in modern LLMs and analyze how intermediate representations evolve during training. We introduce a metric to quantify a model's intrinsic suitability for early-exit and propose a benchmark for researchers to explore the potential early-exit benefits on different models and workloads. Our results show a diminishing trend in early-exit effectiveness across newer model generations. We further find that dense transformers generally offer greater early-exit potential than Mixture-of-Experts and State Space Models. In addition, larger models, particularly those with more than 20 billion parameters, and base pretrained models without specialized tuning tend to exhibit higher early-exit potential.

Figures (10)

• Figure 1: Layer-wise early-exit decoding in LLMs.
• Figure 2: The trend of relative early-exit scores (§\ref{['subsec:metrics']}) in recent LLM and models specifically tuned for early-exit, compared to Llama2-7B. We explain the model selection details in Appendix \ref{['appendix:evaluate_ee_score']}.
• Figure 3: Workflow illustration of the paper.
• Figure 4: The layer-to-final similarity results of eight different models aggregated across four datasets.
• Figure 5: The accuracy and skip ratio of different exit strategies and thresholds.