I am a final-year Artificial Intelligence student at the University of Science, VNU-HCM (HCMUS), and an AI Engineer at Realtime Robotics, where I work on drone image processing, cross-view geo-localization, and vision-based perception for GNSS-denied navigation. I also collaborate on research with Prof. Long Tran-Thanh at the University of Warwick and Prof. The Anh Han at Teesside University, working on physics-informed machine learning, continual and few-shot learning, computer vision, and the game-theoretic analysis of LLM agents.

My interest in physics-informed machine learning grows out of a long-standing fascination with the Earth and sky — atmospheric processes, geological systems, and astronomy have drawn my curiosity since childhood, and I find deep motivation in building models that respect the physical laws governing these systems rather than treating them as black boxes. I am seeking PhD opportunities to advance physics-informed machine learning for scientific discovery.

Outside of research, I write about the things I’m learning on my blog — language learning, astronomy, meteorology and geology, and AI.

📄 Download CV

🔬 Research Interests

Physics-Informed Machine Learning  |  AI for Science  |  Continual & Few-Shot Learning  |  Game-Theoretic Analysis of LLM Agents

📰 Updates

• 2026.08:  📝 Posted a new preprint, “Humans Are More Diverse: Frontier LLMs Show Extreme Policies in Idealised AI Development Races,” with Prof. The Anh Han and collaborators.
• 2026.05:  📝 Posted three preprints: MIST (streaming decision trees for online class-incremental learning), COMPOSE (compositional generalization in continual few-shot learning), and DISCA (training-free cultural alignment of LLMs) — with Prof. Long Tran-Thanh, and DISCA also with Prof. The Anh Han.
• 2026.05:  📝 Posted a preprint on weather-robust cross-view geo-localization, SkyPart, for GNSS-denied drone navigation.
• 2026.04:  🏆 MEMRES won the FSE-AIware Joint Competition 2026 (Agentic Python Dependency Resolution track).
• 2026.02:  🏆 Team “Prompter x Prompter” placed 1st in both tracks of the NeurIPS 2025 Mouse vs AI Challenge; preprint posted.
• 2026.01:  📝 Posted a preprint on predicting review effort in AI-generated pull requests.
• 2025.12:  📝 Posted two preprints: understanding LLM agent behaviour via game theory (with Prof. The Anh Han), and a multi-modal retrieval approach to image captioning.
• 2025.12:  🏅 Placed 4th in the EVENTA Grand Challenge Track 2 at ACM Multimedia 2025.
• 2025.09:  🎉 Co-authored the SHREC2025/COMPASS paper, published in Computers & Graphics (Q2).
• 2025.04:  🥈 Team “Ai-Yahh” placed 2nd of 18 teams in the SHREC 2025 ROOMELSA challenge.

💼 Experience

📝 Publications

Newest first. Unless noted otherwise, entries below are arXiv preprints currently under review.

1. Humans Are More Diverse: Frontier LLMs Show Extreme Policies in Idealised AI Development Races
   Phu-Hoa Pham, Duy-Minh Dao-Sy, Trung-Kiet Huynh, Phu-Quy Nguyen-Lam, Chi-Nguyen Tran, Minh-Trung Le, Phong-Hao Le, Dinh-Nam Nguyen, Thien-Ky Nguyen-Dong, Elias Fernandez Domingos, Le Hong Trang, The Anh Han
   arXiv, 2026  ·  first author

2. MIST: Reliable Streaming Decision Trees for Online Class-Incremental Learning via McDiarmid Bound
   Phu-Hoa Pham, Chi-Nguyen Tran, Nguyen Lam Phu Quy, Dao Sy Duy Minh, Huynh Trung Kiet, Long Tran-Thanh
   arXiv, 2026  ·  co-first author

3. Unlocking Compositional Generalization in Continual Few-Shot Learning (COMPOSE)
   Phu-Quy Nguyen-Lam, Phu-Hoa Pham, Dao Sy Duy Minh, Chi-Nguyen Tran, Huynh Trung Kiet, Long Tran-Thanh
   arXiv, 2026  ·  co-first author

4. Weather-Robust Cross-View Geo-Localization via Prototype-Based Semantic Part Discovery (SkyPart)
   Chi-Nguyen Tran, Dao Sy Duy Minh, Huynh Trung Kiet, Nguyen Lam Phu Quy, Phu-Hoa Pham, Long Tran-Thanh
   arXiv, 2026

5. Training-Free Cultural Alignment of Large Language Models via Persona Disagreement (DISCA)
   Huynh Trung Kiet, Dao Sy Duy Minh, Tuan Nguyen, Chi-Nguyen Tran, Phu-Hoa Pham, Nguyen Lam Phu Quy, The Anh Han, Long Tran-Thanh
   arXiv, 2026

6. MEMRES: A Memory-Augmented Resolver with Confidence Cascade for Agentic Python Dependency Resolution
   Dao Sy Duy Minh, Tran Chi Nguyen, Trung Kiet Huynh, Phu-Hoa Pham, Nguyen Lam Phu Quy, Vu Nguyen
   arXiv, 2026  ·  Winner, FSE-AIware Joint Competition 2026

7. Navigating Simply, Aligning Deeply: Winning Solutions for Mouse vs. AI 2025
   Phu-Hoa Pham, Chi-Nguyen Tran, Dao Sy Duy Minh, Nguyen Lam Phu Quy, Huynh Trung Kiet
   arXiv, 2026  ·  first author, Top-1 (Both Tracks), NeurIPS 2025 Mouse vs AI Challenge

8. Early-Stage Prediction of Review Effort in AI-Generated Pull Requests
   Dao Sy Duy Minh, Huynh Trung Kiet, Nguyen Lam Phu Quy, Phu-Hoa Pham, Tran Chi Nguyen, Nguyen Dinh Ha Duong, Truong Bao Tran
   arXiv, 2026

9. Understanding LLM Agent Behaviours via Game Theory: Strategy Recognition, Biases and Multi-Agent Dynamics
   Trung-Kiet Huynh, Duy-Minh Dao-Sy, Thanh-Bang Cao, Phong-Hao Le, Hong-Dan Nguyen, Phu-Quy Nguyen-Lam, Minh-Luan Nguyen-Vo, Hong-Phat Pham, Phu-Hoa Pham, Thien-Kim Than, Chi-Nguyen Tran, Huy Tran, Gia-Thoai Tran-Le, Alessio Buscemi, Le Hong Trang, The Anh Han
   arXiv, 2025

10. Beyond Vision: Contextually Enriched Image Captioning with Multi-Modal Retrieval
    Nguyen Lam Phu Quy, Phu-Hoa Pham, Tran Chi Nguyen, Dao Sy Duy Minh, Nguyen Hoang Minh Ngoc, Huynh Trung Kiet
    arXiv, 2025

11. Leveraging Lightweight Entity Extraction for Scalable Event-Based Image Retrieval
    Dao Sy Duy Minh, Huynh Trung Kiet, Nguyen Lam Phu Quy, Phu-Hoa Pham, Tran Chi Nguyen
    arXiv, 2025  ·  4th place, EVENTA Grand Challenge Track 2 @ ACM Multimedia 2025

12. SHREC2025: Intelligent 3D Room Design — 3D Shape Retrieval (COMPASS)
    Phu-Hoa Pham (Team Lead), et al. — Team “Ai-Yahh,” 2nd place of 18 teams
    Computers & Graphics (Q2 Journal), 2025  ·  [GitHub]

🎖 Honors and Awards

• Winner — FSE-AIware Joint Competition 2026, Track: “Agentic Based Python Dependency Resolution,” Team “MEMRES.”
• Top-1 (Both Tracks) — NeurIPS 2025: Mouse vs AI, Robust Visual Foraging Challenge, Team “Prompter x Prompter.”
• Silver Medal — NVIDIA Nemotron Model Reasoning Challenge (Kaggle), ranked 99/4,182 teams (top ~2.4%), Team “Continuer.”
• 2nd Place — SHREC2025 (3D Shape Retrieval Contest), Track: “Intelligent 3D Room Design,” April 2025.
• 4th Place — EVENTA Grand Challenge Track 2, ACM Multimedia 2025.
• Top 2 — TRACS @ WASP (Third Workshop on AI for Scientific Publications), IJCNLP-AACL 2025.
• Second Prize — Provincial Excellent Student Contest in Mathematics, Dak Lak Province.

📖 Education

• 2023 – Present, Bachelor of Artificial Intelligence, VNU-HCM University of Science (HCMUS), Ho Chi Minh City, Vietnam
  Current GPA: 9.574/10.0 (3.94/4.0 US equivalent) — Ranked in the top 3% of cohort in Year 1 (2023–2024) and Year 2 (2024–2025).
  Full transcript →

📚 Academic Services

Reviewer

• 3rd Workshop on New Trends in AI-Generated Media and Security, CVPR 2026

💻 Skills

Programming Languages: Python, C/C++, Java
Deep Learning & AI: PyTorch, TensorFlow, Scikit-learn, NumPy, Pandas, Matplotlib
Computer Vision & 3D: OpenCV, PyTorch3D, Open3D, Point Cloud Processing, 3D Data Visualization
Developer Tools: Git, GitHub, VS Code, Visual Studio, PyCharm, CLion
Documentation: LaTeX, Markdown
Languages: English (IELTS 7.0), Vietnamese (Native)