About Me

I am a fourth-year Ph.D. candidate at KAUST in Saudi Arabia. I am part of Vision-CAIR group led by Prof. Mohamed Elhoseiny. My research lies in enhancing foundation models with emerging abilities and properties, including but not limited to:

• Agentic Reasoning. Bootstrapping the agentic ability of VLMs at both training and inference ime.
• Alignment. Training LLMs/VLMs to follow human instructions.
• Mitigating Forgetting. Preserving general knowledge while adapting VLMs to specialized tasks.

Education

PhD Candidate in Computer Science, King Abdullah University of Science and Technology (KAUST)
MA in Applied Mathematics and Computational Science, 2021, University of Pennsylvania
BSc in Mathematics and Applied Mathematics, 2019, Beijing Normal University

News

• [Pinned to top] I am actively seeking a full-time position in 2026. I am happy to engage in discussions regarding potential opportunities!
• [05/2026] Our dTRPO paper on policy optimization for diffusion large language models is accepted to ICML' 26! Code and project page are available here and here.
• [10/2025] Check our paper in gym for VLM agentic training!
• [10/2025] I am awarded the KAUST Dean's List Award by my university!
• [09/2025] Our paper, Vgent, about long video understanding is accepted to NeurIPS '25 as spotlight presentation!
• [08/2025] Started my internship with Meta AI, working on alignment of dLLM.
• [03/2025] Our ICLR '25 (spotlight) work BFPO is out on arxiv! Please find our code here
• [02/2025] Our papers, BFPO about safety alignment and QKT about collaborative learning, are accepted to ICLR '25! See you in Singapore!
• [10/2024] Started my internship with Samsung Research America, working on model merging.
• [06/2024] Our paper, IWMS (Label Delay in CL), is accepted to NeurIPS '24! Code and applications are here !
• [06/2024] Our paper, SPU, is accepted to CVPR '24! Code and applications are here!
• [01/2024] One paper, DietCL, about constrained CL is accepted to ICLR '24!
• [07/2023] Started my internship at Oxford, working on safety alignment.
• [07/2023] One paper is accepted to ICCV '23!
• [04/2023] One paper is accepted to ICML'23!
• [01/2022] Started my Ph.D at KAUST.
• [12/2021] Defended my master thesis, titled Factorized lifelong machine learning on non-stationary tasks: An algorithm and analysis.

Selected Publications [Google Scholar]

Services

Reviewers

• ICLR, NeurIPS, CVPR, ICCV, TPAMI, ICML
• CLAI Unconf, WECIA, ICCVCLVL

Teaching

• Teaching Assistant of CS 326, Low Resource Deep Learning, KAUST

Contact

Address: 4335 WS - 11, Building 1 - The Spine, KAUST 23955, Saudi Arabia
Email: wenxuan.zhang@kaust.edu.sa