Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation

Researcher focused on AGI, multimodal models, and reasoning. Coauthor of Janus and JanusFlow.

PhD student at Tsinghua University focusing on multimodal large language models, reasoning, and reinforcement learning.

PhD student at Tsinghua University focusing on LLM reasoning, RLHF, and multimodal large language models; research intern at DeepSeek.

DeepSeek researcher focused on NLP, code intelligence, and LLM reasoning, with public work spanning DeepSeek-Coder, DeepSeekMath, DeepSeek-V2, DeepSeek-V3, and DeepSeek-R1.

Researcher at DeepSeek AI working on decision-making and post-training for large language models.

Researcher at DeepSeek and a first-year computer science PhD student at the University of Science and Technology of China; works on multimodal reasoning and world models; coauthor of Janus.

Research scientist at DeepSeek and PhD student at the University of Illinois Urbana-Champaign working on multimodal foundation models, large language models, and embodied AI.

Xinlong Wang is a researcher working across computer vision, embodied AI, robotics, and machine learning. Public profiles link him to OpenGVLab and Shanghai AI Laboratory, and he is a coauthor of DeepSeek-VL2.

Research intern at DeepSeek and master's student at Tsinghua University working on large language models, multimodal models, and reinforcement learning.

Member of Technical Staff at DeepSeek.

Research scientist at DeepSeek with public GitHub work on language models and AI systems.

First-year PhD student at Shanghai Jiao Tong University focused on multimodal large language models, text-to-image generation, and image/video generation; coauthor of DeepSeek-VL2.

Yonggang Zhang is a researcher whose public OpenReview profile includes the DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding paper.

Research intern at DeepSeek and PhD student at Stanford University working on generative vision-language models, large language models, and large-scale training.

PhD student at the University of Science and Technology of China focused on machine learning and multimodal understanding and generation; coauthor of Janus.

Research scientist at ByteDance Seed focused on multimodal representation learning, self-supervised learning, and diffusion models; coauthor of Janus and JanusFlow.

PhD student at Shanghai Jiao Tong University working on multimodal large language models and image understanding and generation; coauthor of Janus.

PhD student at The University of Hong Kong focused on large multimodal models and data-centric AI, especially multimodal understanding and generation; coauthor of Janus.

Research assistant at The University of Hong Kong focused on multimodal reasoning and generation, large language models, and embodied AI; coauthor of Janus.

PhD student at The University of Hong Kong focused on multimodal large language models, image and video understanding, generation, and editing; coauthor of Janus.

PhD student at The University of Hong Kong focused on large multimodal models, image and video generation, and multimodal understanding; coauthor of Janus.

Researcher affiliated with DeepSeek-AI and co-author of the Nature paper introducing DeepSeek-R1.