Mohamed S. Abdelfattah

FlashDLM: Accelerating Diffusion Language Model Inference via Efficient KV Caching and Guided Diffusion

2025 · arXiv (Cornell University)

Diffusion language models offer parallel token generation and inherent bidirectionality, promising more efficient and powerful sequence modeling compared to autoregressive approaches. However, state-of-the-art diffusion models (e.g., Dream 7B, LLaDA 8B) suffer from slow inference. While …