About Me

I am a PhD student at the University of Maryland College Park working with Prof. Furong Huang. Previously, I earned a Master's degree from the Singapore University of Technology and Design advised by Prof. Ngai-Man Cheung and a Bachelor from the University of Information Technology - Vietnam National University Ho Chi Minh City working with Dr. Tiep Nguyen and Prof. Tam Nguyen.

My research focuses on data privacy and reliable AI for multimodal and agentic systems.

Email Google Scholar GitHub LinkedIn CV

News

June, 2026 1 paper accepted to EMNLP 2026:
May, 2026 SoundnessBench is now on arXiv.
March, 2026 2 papers accepted to CVPR 2026: 1 Main track paper and 1 Findings track paper
Sep, 2025 Started my PhD in CS at University of Maryland, College Park
Previous news
Sep, 2024 One paper accepted to NeurIPS 2024. Embedding dimension is key for ViT Out-of-Distribution generalization
Jul, 2024 One paper accepted to ECCV 2024. Skip connection reinforces adversary
Mar, 2024 One paper accepted to CVPR 2024. Transfer learning prevents data leakage
Jan, 2024 Started my Master of Engineering at SUTD

Education

Experience

Selected Publications

CVPR 3 ECCV 1 NeurIPS 1 EMNLP 1
Memory Makes The Poison teaser
Memory Makes The Poison: Over Memorization Drives Visual Poisoning in LVLMs

Sy-Tuyen Ho, Y.R.A. Epa, K. Liu, X. Jiang, A. Kot, F. Huang, N.M. Cheung

Preprint
Data Poisoning LVLM
Paper Logo Paper GitHub Logo Code Project
Model Inversion Evaluation teaser
Revisiting Model Inversion Evaluation: From Misleading Standards to Reliable Privacy Assessment

Sy-Tuyen Ho, K.J. Hao, N.B. Nguyen, A. Binder, N.M. Cheung

CVPR 2026 Findings
Data Privacy Model Inversion LVLM
ArXiv Logo Paper HF Logo Dataset GitHub Logo Code Project
Vision Transformer NAS teaser
Vision Transformer Neural Architecture Search for Out-of-Distribution Generalization

Sy-Tuyen Ho*, T.V. Vo*, S. Ebrahimkhani*, N.M. Cheung

NeurIPS 2024
OoD Generalization ViT
ArXiv Logo Paper HF Logo Dataset GitHub Logo Code Project
Skip Connections teaser
On the Vulnerability of Skip Connections to Model Inversion Attacks

K.J. Hao*, Sy-Tuyen Ho*, N.B. Nguyen, N.M. Cheung

ECCV 2024
Data Privacy Model Inversion
ArXiv Logo Paper GitHub Logo Code Project
Model Inversion Robustness teaser
Model Inversion Robustness: Can Transfer Learning Help?

Sy-Tuyen Ho, K.J. Hao, K. Chandrasegaran, N.B. Nguyen, N.M. Cheung

CVPR 2024
Data Privacy Model Inversion
ArXiv Logo Paper GitHub Logo Code Project
SoundnessBench teaser
SoundnessBench: Can Your AI Scientist Really Tell Good Research Ideas from Bad Ones?

Sy-Tuyen Ho, Minghui Liu, Huy Nghiem, Furong Huang

Preprint
AI Scientist LLM
ArXiv Logo Paper HF Logo Dataset GitHub Logo Code Project
Trait-space monitoring results across language models
Trait-space Monitoring for Emerent Misalignment During Supervised Finetuning

Huy Nghiem, Sy‑Tuyen Ho, Sarah Wiegreffe, Hal Daumé III

Preprint
Safety Alignment LLM
Paper Logo Paper
Agentic Critical Training teaser
Agentic Critical Training

Weize Liu, Minghui Liu, Sy-Tuyen Ho, Souradip Chakraborty, Xiyao Wang, Furong Huang

Preprint
Agentic AI LLM RL
Paper Logo Paper
Bias in the Tails evaluation results
Bias in the Tails: How Name-conditioned Evaluative Framing in Resume Summaries Destabilizes LLM-based Hiring

Huy Nghiem, Phuong-Anh Nguyen-Le, Sy-Tuyen Ho, Hal Daume III

EMNLP 2026
AI for Social Good LLM
Paper Logo Paper
Model inversion strategies for vision-language models
Model Inversion Attacks on Vision-Language Models: Do They Leak What They Learn?

N.B. Nguyen, Sy-Tuyen Ho, K.J. Hao, N.M. Cheung

CVPR 2026
Data Privacy Model Inversion LVLM
ArXiv Logo Paper GitHub Logo Code Project

Personal Interests

Ping Pong Cooking Hiking

Beyond research, I'm always up for a friendly ping pong match, experimenting with new recipes in the kitchen or exploring scenic trails on a hike!