Gia Binh Nguyen

Gia Binh Nguyen
Email: binhng2803@gmail.com · Phone number: (+84) 911 212 262
Research Interests: Efficient AI, Model Compression, Multimodal Models

I recently graduated with a B.Sc. in Artificial Intelligence from FPT University. I am currently an AI Research Resident at the Center for Artificial Intelligence Research (CAIR), VinUni, where I am advised by Dr. Duy H. M. Nguyen (IMPRS-IS) , and working on vision-language-action models for robotic manipulation and efficient model.

My research passion lies in Efficient AI — making powerful models faster, smaller, and more deployable in the real world.

I am actively seeking Master's or Ph.D. scholarship opportunities to dive deeper into this space. Feel free to reach out!

Education

B.Sc. in Artificial Intelligence 2022 – 2026
FPT University, Hanoi, Vietnam
Graduated with honors. GPA: 8.59/10

Work Experience

AI Research Resident Apr 2025 – Present
AI Engineer Freelancer May 2024 – Apr2025
Upwork · Remote
AI Engineer Jun 2024 – Feb 2025
BSM Labs · Hanoi, Vietnam
AI Intern Mar 2024 – Jun 2024
Aimesoft · Hanoi, Vietnam

Publications

FOCA thumbnail
ICML 2026
FOCA: Future-Oriented Conditioning for Data-Efficient Vision-Language-Action Adaptation
Duc M. Nguyen*, Nghiem T. Diep*, Binh G. Nguyen*, Trong-Bao Ho, Doanh Le, Tan Nguyen, Thien-Loc Ha, Tran Nhiem, Bao Thach, Nhat Tran, Binh G. Nguyen*, Duc M. Nguyen*, Nghiem T. Diep*, Trong-Bao Ho, Doanh Le, Tan Nguyen, Thien-Loc Ha, Tran Nhiem, Bao Thach, Nhat Tran, Tuan Anh Tran, Artur Habuda, Philip Lund Moeller, Tran Nguyen Le, Daniel Sonntag, Mathias Niepert, Khoa Doan, Vu Duong, Hung Ngo, Minh Vu, Duy MH Nguyen, An Thai Le, Vien Ngo
40th International Conference on Machine Learning (ICML) · 2026
*Co-first authors
CLP thumbnail
RSS Workshop
Finetuning Vision‑Language‑Action Models Requires Fewer Layers Than You Think
Binh G. Nguyen*, Trong-Bao Ho, Thien-Loc Ha, Khoa Vo, Philip Lund Møller, Quang Tan Nguyen, Long Dinh, Tuan Quan Dam, Vu N. Duong, Tung Minh Luu, Trung Le, Binh G. Nguyen*, Trong-Bao Ho, Thien-Loc Ha, Khoa Vo, Philip Lund Møller, Quang Tan Nguyen, Long Dinh, Tuan Quan Dam, Vu N. Duong, Tung Minh Luu, Trung Le, Tran Nguyen Le, Minh Nhat Vu, An Thai Le, Ngan Le, Daniel Sonntag, James Zou, Jan Peters, Duy Minh Nguyen Ho, Vien Anh Ngo
Under Review · 2026 · RSS Workshop on Diffusion for Robot Learning
PAINT thumbnail
CoRL 2026
Start Right, Arrive Right: Asynchronous Execution via Initial Noise Selection
Trong-Bao Ho*, Quang Tan Nguyen*, Thien-Loc Ha*, Binh G. Nguyen, Viet-Thanh Nguyen, Long Dinh, Minh Nhat VU, Duy Minh Ho Nguyen, An Thai Le, Vien Anh Ngo
10th Conference on Robot Learning (CoRL) · 2026
Equivla thumbnail
RSS Workshop
EquiVLA: A General Framework for Rotationally Equivariant Vision–Language–Action Models
Thien-Loc Ha*, Quang Tan Nguyen*, Trong-Bao Ho*, Long Dinh, Minh Duc Nguyen, Binh G. Nguyen, Pham Tri Quang, Minh Nhat VU, Duy Minh Ho Nguyen, An Thai Le, Vien Anh Ngo
Under Review · 2026 · RSS Workshop on Diffusion for Robot Learning
Preference thumbnail
CoRL 2026
Progress-Guided Preference Generation for Aligning Generative Control Policies
Tung Minh Luu*, Tri Ton, Minh Duc Nguyen, Thanh Xuan Nguyen, Nguyen Gia Binh, Quang Tan Nguyen, Vu N. Duong, Daniel Sonntag, Duy Minh Ho Nguyen, Vien Anh Ngo, Chang D. Yoo
10th Conference on Robot Learning (CoRL) · 2026
self thumbnail
Under Review
Self-Improving VLA Policies: Selected Diffusion Noise for Spurious-Robust Action Smoothing
Minh Duc Nguyen*, Bao-Ngoc Dao, Tung Minh Luu, Nguyen Gia Binh, Vinh Tong, Anji Liu, Vu N. Duong, Dung D. Le, Daniel Sonntag, Minh Duc Nguyen*, Bao-Ngoc Dao, Tung Minh Luu, Nguyen Gia Binh, Vinh Tong, Anji Liu, Vu N. Duong, Dung D. Le, Daniel Sonntag, Trung Le, Ngan Le, Jan Peters, An Thai Le, Minh Nhat VU, Mathias Niepert, Khoa D Doan, Duy Minh Ho Nguyen, Vien Anh Ngo
Under Review · 2026

Open-source Contributions

A curated toolbox of data-conversion utilities for LeRobot. Contributed the RoboCasa → LeRobot dataset conversion pipeline.
Repository associated with paper titled "CLP: Finetuning Vision-Language-Action Models Requires Fewer Layers Than You Think"

Honors & Awards

Champion 2024
Funix GenAI Hackathon: MindGPT - Mindmap generation web-app
Runner-up 2024
FPT x NRC Hackathon: Auto data analysis web-app
Finalist 2024
Top 8/70 Teams — Pixta AI Hackathon Face Analysis Challenge
FPTU 100% Scholarship 2022 – 2026
FPT University — covers full 4-year tuition fee; highest Logic Math score (86/90)
Honorable Student Award 2022 – 2026
FPT University — all semesters