EB
Emad Barsoum
cs.CVcs.AIcs.CLcs.LGcs.DCcs.HCcs.PFcs.ROcs.NE
On Valency
published · living versionsW_zdks3phy·v1 · currentpublished
AMDKernelVault: Large-Scale Datasets and Agentic Training for AMD GPU Kernel Optimization
with Ji Liu, Saptarshi Majumder, Yiqing Huang, Wenwen Ouyang +12
1 version
Preprints & journals
86 papers in the corpus · 2016–2026SliMOO: Interpretable Multi-Objective Evolutionary Search for LLM Depth Pruning2610.27812v1 · Guanchen Li, Yixing Xu, Xuanwu Yin et al.2026 · 0 citationsarXiv
Beyond the Trial-and-Error Loop: Hybrid Projection and Automated Tuning for Distributed Training2610.19423v1 · Anshu Raina, Peyman Razaghi, Yuankai Chen et al.2026 · 0 citationsarXiv
Zebra3D: Efficient Hybrid Foundation Models For Scalable 3D Reconstruction2610.12727v1 · Utkarsh Saxena, Ruijing Yang, Mehdi Rezagholizadeh et al.2026 · 0 citationsarXiv
CORAL-KV: Compact Cross-Layer Prediction and Residual Coding for Autoregressive Video KV Caches2610.12514v1 · Albert Ge, Akash Haridas, Utkarsh Saxena et al.2026 · 0 citationsarXiv
CompassPlay: Rewarding the Proposer for Where It Moves the Solver2609.32228v1 · Sophia Xiao Pu, Ximeng Sun, Jiang Liu et al.2026 · 0 citationsarXiv
AMDKernelVault: Large-Scale Datasets and Agentic Training for AMD GPU Kernel Optimization2609.12471v1 · Ji Liu, Saptarshi Majumder, Yiqing Huang et al.2026 · 0 citationsarXivon Valency
Instella-MoE Technical Report2609.00791v1 · Jiang Liu, Sudhanshu Ranjan, Prakamya Mishra et al.2026 · 0 citationsarXiv
AsmEvo: Agentic Assembly-Level Optimization of AMD GPU Kernels with Functional Equivalence Verification2608.20711v1 · Ji Liu, Puyuan Yang, Rongzhang Zheng et al.2026 · 0 citationsarXiv
VEGAS: Human-Aligned Video Caption Evaluation via Gaze2607.08489v1 · Shenghui Chen, Po-han Li, Ximeng Sun et al.2026 · 0 citationsarXiv
Arbor: Tree Search as a Cognition Layer for Autonomous Agents2606.12563v1 · Neha Prakriya, Chaojun Hou, Zheng Gong et al.2026 · 0 citationsarXiv
Pause and Think: A Dataset and Benchmark for Video-Grounded Assistive Action Suggestion2606.00616v3 · Shivam Singh, Saptarshi Majumder, Pratik Prabhanjan Brahma et al.2026 · 0 citationsarXiv
Efficient Diffusion LLMs via Temporal-Spatial Parallel Decoding and Confidence Extrapolation2605.30753v1 · Zekai Li, Ji Liu, Yiqing Huang et al.2026 · 0 citationsarXiv
PR2: Predictive Routing Replay for MoE-Based LLM Reinforcement Learning2606.00395v2 · Daize Dong, Junlin Chen, Haolong Jia et al.2026 · 0 citationsarXiv
Beyond the Target: From Imitation to Collaboration in Speculative Decoding2605.24793v1 · Jinze Li, Yixing Xu, Guanchen Li et al.2026 · 0 citationsarXiv
AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents2605.16819v1 · Sharareh Younesian, Wenwen Ouyang, Sina Rafati et al.2026 · 0 citationsarXiv
Theory-optimal Quantization Based on Flatness2605.18800v1 · Xiusheng Huang, Zhe Li, Xuanwu Yin et al.2026 · 0 citationsarXiv
PARD-2: Target-Aligned Parallel Draft Model for Dual-Mode Speculative Decoding2605.08632v1 · Zihao An, Taichi Liu, Ziqiong Liu et al.2026 · 0 citationsarXiv
Long-Context Aware Upcycling: A New Frontier for Hybrid LLM Scaling2604.24715v1 · Parsa Ashrafi Fashi, Utkarsh Saxena, Mehdi Rezagholizadeh et al.2026 · 0 citationsarXiv
Ego-InBetween: Generating Object State Transitions in Ego-Centric Videos2604.17749v1 · Mengmeng Ge, Takashi Isobe, Xu Jia et al.2026 · 0 citationsarXiv
DiffSparse: Accelerating Diffusion Transformers with Learned Token Sparsity2604.03674v1 · Haowei Zhu, Ji Liu, Ziqiong Liu et al.2026 · 0 citationsarXiv
VideoSeek: Long-Horizon Video Agent with Tool-Guided Seeking2603.20185v1 · Jingyang Lin, Jialian Wu, Jiang Liu et al.2026 · 0 citationsarXiv
DC-DiT: Adaptive Compute and Elastic Inference for Visual Generation via Dynamic Chunking2603.06351v2 · Akash Haridas, Utkarsh Saxena, Parsa Ashrafi Fashi et al.2026 · 0 citationsarXiv
DUET-VLM: Dual stage Unified Efficient Token reduction for VLM Training and Inference2602.18846v2 · Aditya Kumar Singh, Hitesh Kandala, Pratik Prabhanjan Brahma et al.2026 · 0 citationsarXiv
AdaptEvolve: Improving Efficiency of Evolutionary AI Agents through Adaptive Model Selection2602.11931v2 · Pretam Ray, Pratik Prabhanjan Brahma, Zicheng Liu et al.2026 · 0 citationsarXiv
TermiGen: High-Fidelity Environment and Robust Trajectory Synthesis for Terminal Agents2602.07274v1 · Kaijie Zhu, Yuzhou Nie, Yijiang Li et al.2026 · 1 citationarXiv
Learnable Permutation for Structured Sparsity on Transformer Models2601.22980v1 · Zekai Li, Ji Liu, Guanchen Li et al.2026 · 0 citationsarXiv
DiffBench Meets DiffAgent: End-to-End LLM-Driven Diffusion Acceleration Code Generation2601.03178v1 · Jiajun jiao, Haowei Zhu, Puyuan Yang et al.2026 · 0 citationsarXiv
CD4LM: Consistency Distillation and aDaptive Decoding for Diffusion Language Models2601.02236v1 · Yihao Liang, Ze Wang, Hao Chen et al.2026 · 0 citationsarXiv
Dual LoRA: Enhancing LoRA with Magnitude and Direction Updates2512.03402v4 · Yixing Xu, Chao Li, Xuanwu Yin et al.2025 · 0 citationsarXiv
Training-Free Loosely Speculative Decoding: Accepting Semantically Correct Drafts Beyond Exact Match2511.22972v3 · Jinze Li, Yixing Xu, Guanchen Li et al.2025 · 0 citationsarXiv
CaptionQA: Is Your Caption as Useful as the Image Itself?2511.21025v2 · Shijia Yang, Yunong Liu, Bohan Zhai et al.2025 · 0 citationsarXiv
Training Foundation Models on a Full-Stack AMD Platform: Compute, Networking, and System Design2511.17127v2 · Quentin Anthony, Yury Tokpanov, Skyler Szot et al.2025 · 0 citationsarXiv
FarSkip-Collective: Unhobbling Blocking Communication in Mixture of Experts Models2511.11505v3 · Yonatan Dukler, Guihong Li, Deval Shah et al.2025 · 0 citationsarXiv
Instella: Fully Open Language Models with Stellar Performance2511.10628v2 · Jiang Liu, Jialian Wu, Xiaodong Yu et al.2025 · 0 citationsarXiv
Learning from Online Videos at Inference Time for Computer-Use Agents2511.04137v1 · Yujian Liu, Ze Wang, Hao Chen et al.2025 · 0 citationsarXiv
E-MMDiT: Revisiting Multimodal Diffusion Transformer Design for Fast Image Synthesis under Limited Resources2510.27135v1 · Tong Shen, Jingai Yu, Dong Zhou et al.2025 · 0 citationsarXiv
DRIFT: Transferring Reasoning Priors for Efficient MLLM Fine-Tuning2510.15050v2 · Chao Huang, Zeliang Zhang, Jiang Liu et al.2025 · 0 citationsarXiv
XModBench: Benchmarking Cross-Modal Capabilities and Consistency in Omni-Language Models2510.15148v2 · Xingrui Wang, Jiang Liu, Chao Huang et al.2025 · 1 citationPublished as a conference paper at ICLR 2026
ImageDoctor: Diagnosing Text-to-Image Generation via Grounded Image Reasoning2510.01010v1 · Yuxiang Guo, Jiang Liu, Ze Wang et al.2025 · 0 citationsarXiv
Latent Visual Reasoning2509.24251v2 · Bangzheng Li, Ximeng Sun, Jiang Liu et al.2025 · 0 citationsarXiv
APRIL: Active Partial Rollouts in Reinforcement Learning to Tame Long-tail Generation2509.18521v3 · Yuzhen Zhou, Jiajun Li, Yusheng Su et al.2025 · 0 citationsarXiv
SpecVLM: Fast Speculative Decoding in Vision-Language Models2509.11815v2 · Haiduo Huang, Fuwei Yang, Zhenhua Liu et al.2025 · 0 citationsarXiv
Layout-Conditioned Autoregressive Text-to-Image Generation via Structured Masking2509.12046v3 · Zirui Zheng, Takashi Isobe, Tong Shen et al.2025 · 0 citationsarXiv
SparK: Query-Aware Unstructured Sparsity with Recoverable KV Cache Channel Pruning2508.15212v3 · Huanxuan Liao, Yixing Xu, Shizhu He et al.2025 · 0 citationsarXiv
Geak: Introducing Triton Kernel AI Agent & Evaluation Benchmarks2507.23194v1 · Jianghui Wang, Vinay Joshi, Saptarshi Majumder et al.2025 · 0 citationsarXiv
SAND-Math: Using LLMs to Generate Novel, Difficult and Useful Mathematics Questions and Answers2507.20527v3 · Chaitanya Manem, Pratik Prabhanjan Brahma, Prakamya Mishra et al.2025 · 0 citationsarXiv
Instella-T2I: Pushing the Limits of 1D Discrete Latent Space Image Generation2506.21022v1 · Ze Wang, Hao Chen, Benran Hu et al.2025 · 0 citationsarXiv
Athena: Enhancing Multimodal Reasoning with Data-efficient Process Reward Models2506.09532v5 · Shuai Wang, Zhenhua Liu, Jiaheng Wei et al.2025 · 0 citationsarXiv
TTT-Bench: A Benchmark for Evaluating Reasoning Ability with Simple and Novel Tic-Tac-Toe-style Games2506.10209v1 · Prakamya Mishra, Jiang Liu, Jialian Wu et al.2025 · 0 citationsarXiv
TaDA: Training-free recipe for Decoding with Adaptive KV Cache Compression and Mean-centering2506.04642v1 · Vinay Joshi, Pratik Prabhanjan Brahma, Zicheng Liu et al.2025 · 0 citationsarXiv
Career total: 106 works. 86 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.