DSB: Dynamic Sliding Block Scheduling for Diffusion LLMs

TL;DR

This paper introduces DSB, a dynamic sliding block scheduling method for diffusion LLMs that adapts to semantic difficulty, improving both output quality and inference efficiency without additional training.

Contribution

The paper proposes a novel, training-free dynamic sliding block scheduling approach and a KV-cache mechanism, addressing limitations of fixed schedules in diffusion LLMs.

Findings

Improves generation quality across multiple models and benchmarks.

Enhances inference efficiency without additional training.

Demonstrates robustness and generality of the proposed methods.

Abstract

Diffusion large language models (dLLMs) have emerged as a promising alternative for text generation, distinguished by their native support for parallel decoding. In practice, block inference is crucial for avoiding order misalignment in global bidirectional decoding and improving output quality. However, the widely-used fixed, predefined block (naive) schedule is agnostic to semantic difficulty, making it a suboptimal strategy for both quality and efficiency: it can force premature commitments to uncertain positions while delaying easy positions near block boundaries. In this work, we analyze the limitations of naive block scheduling and disclose the importance of dynamically adapting the schedule to semantic difficulty for reliable and efficient inference. Motivated by this, we propose Dynamic Sliding Block (DSB), a training-free block scheduling method that uses a sliding block with a dynamic size to overcome the rigidity of the naive block. To further improve efficiency, we introduce DSB Cache, a training-free KV-cache mechanism tailored to DSB. Extensive experiments across multiple models and benchmarks demonstrate that DSB, together with DSB Cache, consistently improves both generation quality and inference efficiency for dLLMs. Code is released at https://github.com/lizhuo-luo/DSB.