Accelerate Speculative Decoding with Sparse Computation in Verification
This addresses efficiency issues in speculative decoding for long-context and MoE models, representing an incremental improvement over existing sparsification methods.
The paper tackles the computational bottleneck in the verification stage of speculative decoding for language models by proposing a sparse verification framework that jointly sparsifies attention, FFN, and MoE components, achieving favorable efficiency-accuracy trade-offs across multiple datasets.
Speculative decoding accelerates autoregressive language model inference by verifying multiple draft tokens in parallel. However, the verification stage often becomes the dominant computational bottleneck, especially for long-context inputs and mixture-of-experts (MoE) models. Existing sparsification methods are designed primarily for standard token-by-token autoregressive decoding to remove substantial computational redundancy in LLMs. This work systematically adopts different sparse methods on the verification stage of the speculative decoding and identifies structured redundancy across multiple dimensions. Based on these observations, we propose a sparse verification framework that jointly sparsifies attention, FFN, and MoE components during the verification stage to reduce the dominant computation cost. The framework further incorporates an inter-draft token and inter-layer retrieval reuse strategy to further reduce redundant computation without introducing additional training. Extensive experiments across summarization, question answering, and mathematical reasoning datasets demonstrate that the proposed methods achieve favorable efficiency-accuracy trade-offs, while maintaining stable acceptance length.