Sicheng Feng, Zigeng Chen, Song Wang, Gongfan Fang, Xinchao Wang (National University of Singapore) · Multimodal Speculative Decoding

MMFlash: Faster Speculative Decoding for Vision-Language Models

MMFlash is a speculative decoding framework for vision-language models that uses a diffusion-style drafter to propose token blocks in one forward pass. It adaptively upweights visually dependent tokens that the drafter predicts poorly, and uses sparse attention for long video inputs. The authors report lossless decoding speedups of up to 3.58 times across multimodal benchmarks, outperforming EAGLE-3 and DFlash by up to 195.4% and 17.1% respectively.

Open the source ↗

Similar Research & Safety releases

  • DySENSE

    Kailai Sun, Zijie Gao, Mingyi He, Can Rong, Alok Prakash, Baoshen Guo, Shenhao Wang, Jinhua Zhao (Massachusetts Institute of Technology) · Urban Energy GenAI · Week 48, 2026

  • AllocFlow

    Annan Wang, Yuchen Xie, Yu Cen, Leon Liangyu Chen, Qianyu Julie Zhu, Xin Li, Tianyi Wei, Weisi Lin (Nanyang Technological University) · Flow-Matching Trading · Week 48, 2026

  • MIRIX-2.0: Multi-Agent Memory System

    Wei-Chieh Huang, Weizhi Zhang, Jiayan Nan, Peng Xia, Yu Wang (arXiv) · Agent Memory System · Week 48, 2026

  • A step toward defining the Standard Model on a lattice

    OpenAI · AI-Assisted Proof · Week 48, 2026

  • Memorilla

    Stanford SNAP group (Jure Leskovec), with co-authors from Toronto, Penn and MIT · Latent Memory Module · Week 47, 2026

  • NovaAtom-Lite

    Geodesic Intelligence · All-atom structure prediction · Week 47, 2026

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.