Xubing Ye

Xubing Ye (叶栩冰)

Hi! I'm Xubing Ye. I'm currently an algorithm engineer at ByteDance Seed. I obtained my master's degree from Shenzhen International Graduate School at Tsinghua University in 2026, supervised by Prof. Yansong Tang. I obtained my bachelor's degree from the School of Software Engineering at Tongji University in 2023.

My current research focuses on data engineering for agentic LLMs and MLLMs.

Email: yxb_tongji@163.com / Github / Scholar

Recent Publications

* Indicates Equal Contribution

	VoCo-LLaMA: Towards Vision Compression with Large Language Models Xubing Ye, Yukang Gan, Xiaoke Huang, Yixiao Ge, Yansong Tang IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025 [arXiv] [Code] Compresses hundreds of video tokens into one via attention distillation for efficient long-video understanding.
	ATP-LLaVA: Adaptive Token Pruning for Large Vision Language Models Xubing Ye, Yukang Gan, Yixiao Ge, Xiao-ping Zhang, Yansong Tang IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025 [arXiv] Adaptively prunes visual tokens inside LLM decoder layers with minimal performance loss.
	MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scaling MiroMind Team Technical Report, 2026 [arXiv] [Code] [HF] An open-source deep research agent advancing model, context, and interactive scaling.
	POINTS-Reader: Distillation-Free Adaptation of Vision-Language Models for Document Conversion WeLM Team Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025 [arXiv] [Code] [HF] A distillation-free document conversion model for extracting text, tables, and formulas.
	Language-Aware Vision Transformer for Referring Segmentation Xubing Ye, Zhao Yang, Jiaqi Wang, Yansong Tang, Kai Chen, Hengshuang Zhao, Philip H.S. Torr IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI, IF=20.8), 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)*, 2022 [IEEE] [Conference Version] [Code] A universal referring image and video segmentation framework with language-aware visual encoding.