Profile picture

An Vo

Research Engineer @ Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)

I am now Research Engineer at MBZUAI working with Thamar Solorio. I obtained my MS from KAIST with Anh Totti Nguyen and Daeyoung Kim. During that time, I also worked closely with Mohammad Reza Taesiri. My MS Program was fully funded by Hyundai CMK Global Scholarship (prestigious scholarship for graduate students). Prior to join KAIST, I obtained my BS degree as the valedictorian at Vietnam National University - Ho Chi Minh City (VNU-HCM) in 2023 with Ngoc Hoang Luong.

research

I am broadly interested in Large Language Models (LLMs) and Vision Language Models (VLMs), especially in making them more trustworthy and explainable in edge/hard cases. In my previous life, I worked at the intersection of Evolutionary Computation, Multi-objective Optimization, and AutoML.

Recently, I am the lead author of VLMs are Biased, a paper that was featured on Hacker News (front page; top-5), LinkedIn, Gary Marcus's article (GPT-5: Overdue, overhyped and underwhelming), Lucas Beyer's tweet. In this work, we introduce VLMBias, a benchmark for evaluating visual counting in VLMs. We show that state-of-the-art models (e.g., o3, o4-mini, Gemini 2.5 Pro, Claude 3.7 Sonnet) achieve 100% accuracy counting on images of popular subjects (e.g. knowing that the Adidas logo has 3 stripes and a dog has 4 legs) but are only ~17% accurate in counting in counterfactual images (e.g. counting stripes in a 4-striped Adidas-like logo or counting legs in a 5-legged dog). My work has been used by Google DeepMind and ByteDance and has been accepted at top venues: ICML, AAAI, GECCO, etc.

news

Jan 2026 Starting a new Research Engineer position at MBZUAI. I'm thrilled to begin this next chapter.
Dec 2025 Google DeepMind and ByteDance have used my VLMs are Biased benchmark.
Dec 2025 I successfully defended my MS thesis, relieved and proud to reach this milestone.
Jul 2025 I will attend ICML 2025 at Vancouver to present B-score and VLMs are Biased. Visit our poster to discuss or grab a coffee. Happy to connect with everyone!
Jun 2025 I will serve as a reviewer for AAAI 2026
Jun 2025 One preprint on arXiv: VLMs are Biased
May 2025 One paper has been accepted to ICML 2025
Jan-May 2025 This year, I will serve as a reviewer for multiple AI venues: IEEE CEC 2025, IJCNN 2025, BMVC 2025, and IEEE Transactions on Evolutionary Computation (IF=11.7).
Dec 2024 One paper has been accepted to AAAI 2025
Mar 2024 One paper has been accepted to GEECO 2024
Jan 2024 One paper has been accepted to CEC 2024
Jan 2024 I have been admitted to the MS program at KAIST School of Computing for Spring 2024 and received the KAIST Scholarship. A new journey!
Jan 2024 I have been chosen as a scholar for the Hyundai CMK Global Scholarship 2024, a prestigious scholarship for top Southest Asian students
Jan 2024 This year, I will serve as a reviewer for IEEE CEC 2024
Jul 2023 One paper has been accepted to Informatica
Apr 2023 One paper has been accepted to GECCO 2023 (Late-breaking Abstract)
Oct-Dec 2022 One paper has been accepted to SoICT 2022 and has achieved Best Paper Award

highlights

* indicates equal contribution.

  1. Vision Language Models are Biased
    An Vo*, Khai-Nguyen Nguyen*, Mohammad Reza Taesiri, Vy Tuong Dang, Anh Totti Nguyen†, Daeyoung Kim†
    2025
  2. B-score: Detecting biases in large language models using response history
    An Vo, Mohammad Reza Taesiri, Daeyoung Kim, Anh Totti Nguyen
    International Conference on Machine Learning (ICML), 2025
  3. Efficient Multi-Objective Neural Architecture Search via Pareto Dominance-based Novelty Search
    An Vo, Ngoc Hoang Luong
    Genetic and Evolutionary Computation Conference (GECCO), 2024

recent publications

* indicates equal contribution.

  1. VMMU: A Vietnamese Multitask Multimodal Understanding and Reasoning Benchmark
    Vy Tuong Dang*, An Vo*, Emilio Villa-Cueva, Quang Tau, Duc Dm, Thamar Solorio, Daeyoung Kim
    2026
  2. RADBench: Evaluating Foundation VLMs for Corner Case Description on A Comprehensive Dataset
    Gabriel Manalu, An Vo, Duc Dm, Quang Tau, Dinh Son Nguyen, Minh Son Hoang, Hyeontaek Hwang, Jaesung Lee, Hyojin Choi, Tuan Phong Tran, Daeyoung Kim
    IEEE Intelligent Vehicles Symposium (IV), 2026
  3. Efficient Multi-Objective Neural Architecture Search via Tree Search with Training-Free Metrics
    An Vo, Nhat Minh Le, Ngoc Hoang Luong
    SN Computer Science, 2026
  4. Reference-Based Post-OCR Processing with LLM for Precise Diacritic Text in Historical Document Recognition
    Thao Do, Dinh Phu Tran, An Vo, Daeyoung Kim
    Annual AAAI Conference on Artificial Intelligence (AAAI), 2025
  5. Lightweight multi-objective evolutionary neural architecture search with low-cost proxy metrics
    Ngoc Hoang Luong, Quan Minh Phan, An Vo, Tan Ngoc Pham, Dzung Tri Bui
    Information Sciences, 2024
  6. Zero-Cost Proxy-Based Hierarchical Initialization for Evolutionary Neural Architecture Search
    Nhat Minh Le*, An Vo*, Ngoc Hoang Luong
    IEEE Congress on Evolutionary Computation (CEC), 2024
  7. Accelerating Gene-pool Optimal Mixing Evolutionary Algorithm for Neural Architecture Search with Synaptic Flow
    Khoa Huu Tran, Luc Truong, An Vo, Ngoc Hoang Luong
    Genetic and Evolutionary Computation Conference (GECCO Companion), 2023
  8. Lightweight Multi-Objective and Many-Objective Problem Formulations for Evolutionary Neural Architecture Search with the Training-Free Performance Metric Synaptic Flow
    An Vo, Tan Ngoc Pham, Ngoc Hoang Luong
    Informatica, 2023
  9. Training-Free Multi-Objective and Many-Objective Evolutionary Neural Architecture Search with Synaptic Flow
    An Vo, Tan Ngoc Pham, Van Bich Nguyen, Ngoc Hoang Luong
    International Symposium on Information and Communication Technology (SoICT), 2022
    Best Paper Award

media coverage