AMD

SpecVLM: Fast Speculative Decoding in Vision-Language Models

Haiduo Huang, Fuwei Yang, Zhenhua Liu, Xuanwu Yin, Dong Li, Pengju Ren, Emad Barsoum

机构 * Advanced Micro Devices, Inc.（先进微器件公司） ； Institute of Artificial Intelligence and Robotics（人工智能与机器人研究院） 纠错

Geometric Integration for Neural Control Variates

Daniel Meister, Takahiro Harada

机构 * Advanced Micro Devices, Inc.（先进微器件公司） 纠错

X-EcoMLA: Upcycling Pre-Trained Attention into MLA for Efficient and Extreme KV Compression

Guihong Li, Mehdi Rezagholizadeh, Mingyu Yang, Vikram Appia, Emad Barsoum

机构 * Advanced Micro Devices, Inc.（先进微器件公司） 纠错

A High Efficient and Scalable Obstacle-Avoiding VLSI Global Routing Flow

Junhao Guo, Hongxin Kong, Lang Feng

机构 * School of Integrated Circuits（集成电路学院） ； Sun Yat-sen University（中山大学） ； Advanced Micro Devices, Inc.（先进微系统公司） 纠错

Comments Accepted by ACM TODAES

Batch-Max: Higher LLM Throughput using Larger Batch Sizes and KV Cache Compression

Michael R. Metel, Boxing Chen, Mehdi Rezagholizadeh

机构 * Huawei Noah's Ark Lab（华为诺亚实验室） ； Advanced Micro Devices, Inc.（先进微器件公司） 纠错

Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding

Jinze Li, Yixing Xu, Haiduo Huang, Xuanwu Yin, Dong Li, Edith C. H. Ngai, Emad Barsoum

机构 * Advanced Micro Devices, Inc., Beijing, China（Advanced Micro Devices公司，北京） ； Department of Electrical and Electronic Engineering, The University of Hong Kong（香港大学电子与电气工程系） ； Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University（西安交通大学人工智能与机器人研究所） 纠错

Comments Accepted to the 42nd International Conference on Machine Learning (ICML 2025). Code: https://github.com/AMD-AIG-AIMA/Gumiho

Multi-task parallelism for robust pre-training of graph foundation models on multi-source, multi-fidelity atomistic modeling data

Massimiliano Lupo Pasini, Jong Youl Choi, Pei Zhang, Kshitij Mehta, Rylie Weaver, Ashwin M. Aji, Karl W. Schulz, Jorda Polo, Prasanna Balaprakash

机构 * Oak Ridge National Laboratory（橡树岭国家实验室） ； Advanced Micro Devices Research（先进微器件研究） 纠错

Comments 15 pages, 4 figures, 2 tables

Instella-T2I: Pushing the Limits of 1D Discrete Latent Space Image Generation

Ze Wang, Hao Chen, Benran Hu, Jiang Liu, Ximeng Sun, Jialian Wu, Yusheng Su, Xiaodong Yu, Emad Barsoum, Zicheng Liu

机构 * AMD GenAI（AMD 生成人工智能） 纠错

FINN-GL: Generalized Mixed-Precision Extensions for FPGA-Accelerated LSTMs

Shashwat Khandelwal, Jakoba Petri-Koenig, Thomas B. Preußer, Michaela Blott, Shreejith Shanker

机构 * Reconfigurable Computing Systems Lab（可重构计算系统实验室） ； Advanced Micro Devices (AMD) Research（Advanced Micro Devices (AMD) 研究） 纠错

Comments 9 pages, 6 figures, 5 tables, Accepted for publication in IEEE FPL-2025 (https://2025.fpl.org/)

ReNeg: Learning Negative Embedding with Reward Guidance

Xiaomin Li, Yixuan Liu, Takashi Isobe, Xu Jia, Qinpeng Cui, Dong Zhou, Dong Li, You He, Huchuan Lu, Zhongdao Wang, Emad Barsoum

机构 * Advanced Micro Devices Inc.（先进微器件公司） ； Dalian University of Technology（大连理工大学） ； Tsinghua University（清华大学） 纠错

Comments Code: https://github.com/AMD-AIG-AIMA/ReNeg

TaDA: Training-free recipe for Decoding with Adaptive KV Cache Compression and Mean-centering

Vinay Joshi, Pratik Prabhanjan Brahma, Zicheng Liu, Emad Barsoum

机构 * AMD 纠错

Comments ACL-2025 industry-track accepted

MOVi: Training-free Text-conditioned Multi-Object Video Generation

Aimon Rahman, Jiang Liu, Ze Wang, Ximeng Sun, Jialian Wu, Xiaodong Yu, Yusheng Su, Vishal M. Patel, Zicheng Liu, Emad Barsoum

机构 * Johns Hopkins University（约翰霍普金斯大学） ； Advanced Micro Devices（先进微器件公司） 纠错

Self-Taught Agentic Long Context Understanding

Yufan Zhuang, Xiaodong Yu, Jialian Wu, Ximeng Sun, Ze Wang, Jiang Liu, Yusheng Su, Jingbo Shang, Zicheng Liu, Emad Barsoum

机构 * AMD ； UC San Diego（加州大学圣地亚哥分校） 纠错

Comments Published at ACL 2025 Main Conference

Image-GS: Content-Adaptive Image Representation via 2D Gaussians

Yunxiang Zhang, Bingxuan Li, Alexandr Kuznetsov, Akshay Jindal, Stavros Diolatzis, Kenneth Chen, Anton Sochenov, Anton Kaplanyan, Qi Sun

机构 * New York University（纽约大学） ； Advanced Micro Devices（先进微器件） ； Intel Corporation（英特尔公司） 纠错

Comments ACM SIGGRAPH 2025 Conference Proceedings

FSBench: A Figure Skating Benchmark for Advancing Artistic Sports Understanding

Rong Gao, Xin Liu, Zhuozhao Hu, Bohao Xing, Baiqiang Xia, Zitong Yu, Heikki Kälviäinen

机构 * Lappeenranta-Lahti University of Technology LUT（拉佩兰塔-拉赫蒂技术大学） ； Tianjin University（天津大学） ； AMD Silo AI ； Great Bay University（大湾大学） ； Rensselaer Polytechnic Institute（雷士大学） ； Brno University of Technology（布拉格技术大学） 纠错