About
A little more about me.
I am a Visiting Researcher at Stanford University, working with Joseph C. Wu and Xuekun Wu on agentic AI for life sciences, and a student researcher at Google DeepMind, working with Ming-Hsuan Yang and Liangzhe Yuan on efficient video and vision-language models. Previously, I was a research intern at Nanyang Technological University (with Kwok-Yan Lam and Yongsen Zheng), Meta AI (Llama 4 team), and Snap Research (with Jian Wang), and a researcher at the HCP Lab of Sun Yat-sen University (with Keze Wang and Liang Lin).
My goal is to help build the next generation of AI agents for scientific discovery.
For research and collaboration, write to jusheng@stanford.edu.
Updates
- Two papers accepted to NeurIPS 2026: Self-Rewarded Reasoning (Spotlight) and E0.
- Our work at ECCV 2026: FlashVLM (main conference Poster) and ActPhysCause (LoViF Workshop Oral).
- Joined Stanford University as a Visiting Researcher (terminal-degree requirement waived by the Dean’s Office).
- Hosted the 1st LoViF Challenge on Efficient VLMs at CVPR 2026; challenge report presented as Workshop Oral.
- Three papers accepted to ICML 2026: Failure-Driven Workflow Refinement (Spotlight), SOLAR for Offline MARL, and Great Minds Think Alike.
- Joined Google DeepMind as a student researcher, working with Ming-Hsuan Yang and Liangzhe Yuan.
- Five papers accepted to CVPR 2026, including two Highlights: HybridToken-VLM and Vocabulary Scaling Law.
- Two papers accepted to ICLR 2026: Trading Visual Uncertainties and CoMem.
- Awarded the Snap Global AI Future Leader Scholarship ($30,000).
- Started a research internship at Nanyang Technological University with Kwok-Yan Lam and Yongsen Zheng.
- Three papers accepted to AAAI 2026, including RaCoT (Oral).
- Four papers accepted to NeurIPS 2025: GAM-Agent, MAT-Agent, Tri-MARF, and CF-VLM.
- Two papers accepted to EMNLP 2025, including DrDiff (Oral).
- Selected for the Snap Research Fellowship (top 10 globally, $30,000).
- KABB accepted to ICML 2025.
- Joined Meta AI as a research intern, contributing to the Llama 4 project.
Challenges & Organizing
LoViF: Efficient VLMs for Multimodal Ad Creative Quality Scoring
Lead Organizer · First author of the challenge report
CVPR 2026 · LoViF [Challenge] [Workshop] [Report]
Benchmarking efficient VLMs for multimodal ad quality scoring.
ActPhysCause: Action-Conditioned Physical and Causal World Modeling
Organizer · First author of the baseline report
ECCV 2026 · LoViF [Challenge] [Workshop] [Report]
Action-conditioned video generation for bimanual robot manipulation.
Talks & Presentations
- Spotlight presentation, NeurIPS 2026, Atlanta (upcoming) [Paper]
Self-Rewarded Multimodal Coherent Reasoning Across Diverse Visual Domains · Dec 2026 - Oral, ECCV 2026 LoViF Workshop, Malmö [Paper]
The LoViF 2026 ActPhysCause Challenge: A Baseline Report on Action-Conditioned Video World Models · Sep 2026 - Spotlight presentation, ICML 2026, Seoul [Paper]
Failure-Driven Workflow Refinement · Jul 2026 - Oral, CVPR 2026 LoViF Workshop, Denver [Paper]
The 1st LoViF Challenge on Efficient VLM for Multimodal Ad Creative Quality Scoring · Jun 2026 - Oral, AAAI 2026, Singapore [Paper]
RaCoT: Plug-and-play Contrastive Example Generation Mechanism for Enhanced LLM Reasoning Reliability · Jan 2026 - Oral, EMNLP 2025, Suzhou [Paper]
DrDiff: Dynamic Routing Diffusion with Hierarchical Attention · Nov 2025
Experience
Research
Visiting Researcher · Supervisors: Joseph C. Wu, Xuekun Wu
Exploring agentic AI for biological research and scientific discovery.
Jun 2026 – Present
Student Researcher · Mentors: Ming-Hsuan Yang, Liangzhe Yuan
Efficient video pretraining and vision-language models for high-frame-rate video understanding.
Mar 2026 – Present
Nanyang Technological University
Research Intern · Supervisors: Kwok-Yan Lam, Yongsen Zheng
Robust multi-agent systems under adversarial threats.
Dec 2025 – May 2026
Research Intern · Mentor: Rohan Maheshwari
Contributed to the Llama 4 project as part of the broader team.
Mar 2025 – Aug 2025
Research Collaborator · Mentor: Jian Wang
Controllable video generation with precise prompt-to-effect alignment.
Mar 2024 – Feb 2026
Sun Yat-sen University, HCP Lab
Lab Researcher · Supervisors: Keze Wang, Liang Lin
I am grateful to Prof. Keze Wang for giving me the opportunity to begin my research journey.
Jan 2024 – Dec 2025
Industry
Risk Control Algorithm Expert · Ant Shield
Developed core Ant Shield modules and a dynamic graph neural network pipeline for large-scale fraud and Sybil-attack detection.
2018 – 2023
Professional Service
Conference Reviewing
- 2026: ICLR, ICML, COLM, CVPR, ECCV, ACL (ARR), AAAI, AISTATS
- 2025: NeurIPS, ICLR, ICML, COLM, CVPR, ICCV, WACV, ACL (ARR), EMNLP (ARR), AAAI, AISTATS
- 2024: NeurIPS, ICML, CVPR, ECCV, ACM MM
- 2023: PRCV
Workshops & Challenges
- Lead Organizer, LoViF Challenge on Efficient VLMs at CVPR 2026
- Organizer, LoViF ActPhysCause Challenge at ECCV 2026
Selected Awards
Fellowships & Scholarships
- Snap Global AI Future Leader Scholarship, 2026 ($30,000)
- Snap Research Fellowship, 2025 (top 10 globally, $30,000)
Research & Reviewer Recognition
- Outstanding Reviewer, AAAI, 2026
- Outstanding Research Contribution Award, HCP Lab, Sun Yat-sen University, 2025
- Outstanding Reviewer, ICME and ICASSP, 2025