Fast-dVLM: Efficient Block-Diffusion VLM via Direct Conversion from Autoregressive VLM

• URL: http://arxiv.org/abs/2604.06832v2
• Date: Fri, 10 Apr 2026 04:22:22 GMT
• Title: Fast-dVLM: Efficient Block-Diffusion VLM via Direct Conversion from Autoregressive VLM
• Abstract summary: We present Fast-dVLM, a block-diffusion-based VLM that enables KV-cache-compatible parallel decoding and speculative block decoding for inference acceleration.<n>With SGLang integration and FP8 quantization, Fast-dVLM achieves over 6x end-to-end inference speedup over the AR baseline.
• Score: 58.322826487307765
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Vision-language models (VLMs) predominantly rely on autoregressive decoding, which generates tokens one at a time and fundamentally limits inference throughput. This limitation is especially acute in physical AI scenarios such as robotics and autonomous driving, where VLMs are deployed on edge devices at batch size one, making AR decoding memory-bandwidth-bound and leaving hardware parallelism underutilized. While block-wise discrete diffusion has shown promise for parallel text generation, extending it to VLMs remains challenging due to the need to jointly handle continuous visual representations and discrete text tokens while preserving pretrained multimodal capabilities. We present Fast-dVLM, a block-diffusion-based VLM that enables KV-cache-compatible parallel decoding and speculative block decoding for inference acceleration. We systematically compare two AR-to-diffusion conversion strategies: a two-stage approach that first adapts the LLM backbone with text-only diffusion fine-tuning before multimodal training, and a direct approach that converts the full AR VLM in one stage. Under comparable training budgets, direct conversion proves substantially more efficient by leveraging the already multimodally aligned VLM; we therefore adopt it as our recommended recipe. We introduce a suite of multimodal diffusion adaptations, block size annealing, causal context attention, auto-truncation masking, and vision efficient concatenation, that collectively enable effective block diffusion in the VLM setting. Extensive experiments across 11 multimodal benchmarks show Fast-dVLM matches its autoregressive counterpart in generation quality. With SGLang integration and FP8 quantization, Fast-dVLM achieves over 6x end-to-end inference speedup over the AR baseline.

Related papers

• BARD: Bridging AutoRegressive and Diffusion Vision-Language Models Via Highly Efficient Progressive Block Merging and Stage-Wise Distillation [9.248424980709453]
  We present BARD, a simple and effective bridging framework that converts a pretrained autoregressive VLM into a large-block diffusion VLM.<n>With $leq$ 4.4M data, BARD-VL transfers strong multimodal capability from Qwen3-VL to a large-block dVLM.
  arXiv  Detail & Related papers  (2026-04-15T09:17:38Z)
• VidLaDA: Bidirectional Diffusion Large Language Models for Efficient Video Understanding [52.69880888587866]
  Current Video Large Language Models (Video LLMs) typically encode frames via a encoder vision and employ an autoregressive (AR) LLM for understanding and generation.<n>We propose VidLaDA, a Diffusion Video LLM based on Language Models (DLMs) that leverages bidirectional attention to unlock comprehensive modeling and decode tokens in parallel.<n>Experiments show VidLaDA rivals state-of-the-art AR baselines and outperforms DLM baselines, with MARS-Cache delivering over 12x speedup without compromising accuracy.
  arXiv  Detail & Related papers  (2026-01-25T15:02:01Z)
• WeDLM: Reconciling Diffusion Language Models with Standard Causal Attention for Fast Inference [44.87788417755154]
  We propose WeDLM, a diffusion decoding framework built entirely on standard causal attention.<n>We show that WeDLM preserves the quality of strong AR backbones while delivering substantial speedups.
  arXiv  Detail & Related papers  (2025-12-28T01:25:48Z)
• InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models [49.08289742711585]
  We propose InfiniteVL, a linear-complexity VLM architecture that synergizes sliding window attention (SWA) with Gated DeltaNet.<n>We show that InfiniteVL achieves over 3.6times inference speedup while maintaining constant latency and memory footprint.<n>In streaming video understanding scenarios, it sustains a stable 24 FPS real-time prefill speed while preserving long-term memory cache.
  arXiv  Detail & Related papers  (2025-12-09T17:18:32Z)
• Fast-dLLM v2: Efficient Block-Diffusion LLM [64.38006546510337]
  Fast-dLLM v2 is a block diffusion language model that adapts pretrained AR models into dLLMs for parallel text generation.<n>This represents a 500x reduction in training data compared to full-attention diffusion LLMs such as Dream (580B tokens)
  arXiv  Detail & Related papers  (2025-09-30T14:40:18Z)
• Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding [51.711605076319216]
  Diffusion-based large language models (Diffusion LLMs) have shown promise for non-autoregressive text generation with parallel decoding capabilities.<n>We introduce a novel block-wise approximate KV Cache mechanism tailored for bidirectional diffusion models, enabling cache reuse with negligible performance drop.<n>We propose a confidence-aware parallel decoding strategy that selectively decodes tokens exceeding a confidence threshold, mitigating dependency violations and maintaining generation quality.
  arXiv  Detail & Related papers  (2025-05-28T17:39:15Z)
• MMInference: Accelerating Pre-filling for Long-Context VLMs via Modality-Aware Permutation Sparse Attention [61.025422435235456]
  MMInference is a dynamic sparse attention method that accelerates the prefilling stage for long-context multi-modal inputs.<n>We show that MMInference accelerates the pre-filling stage by up to 8.3x at 1M tokens while maintaining accuracy.
  arXiv  Detail & Related papers  (2025-04-22T17:59:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.