Hi! I'm An 👋

I'm an Assistant Professor at VinUniversity and a Visiting Professor at IAS, TU Darmstadt, where I also lead the AI team at VinRobotics, focused on RL for humanoid locomotion, humanoid VLA architecture and training, and model deployment.

My research lives at the intersection of Robotics and Machine Learning. I try to scale planning and policy learning to settings classical methods struggle with: long horizons, high-dimensional state spaces, large plan sets, and multimodal (often non-convex) solution landscapes, mostly through tensor search (GPU-batched search and optimization over plan tensors) and by pairing algorithmic structure with generative models (diffusion, flow matching). Lately, mostly humanoid loco-manipulation.

On weekends, I write JAX/PyTorch simulators for curved spacetimes (Kerr black holes, the Penrose process, warp-drive energy conditions), because they are beautiful!

Selected papers

Robot Learning & Planning

• Model Tensor Planning, ICLR 2026 / TMLR 2025
• Global Tensor Motion Planning, ICRA 2026 / IEEE RA-L 2025
• Motion Planning Diffusion, IROS 2023 / IEEE T-RO 2025 / AAAI 2026
• Accelerating Motion Planning via Optimal Transport, NeurIPS 2023
• FOCA: Future-Oriented Conditioning for Data-Efficient Vision-Language-Action Adaptation, ICML 2026
• CLOT: Multi-Robot Motion Planning via Collaborative Optimal Transport under STL Tasks, ICRA 2026
• DoublyAware: Dual Planning and Policy Awareness for Temporal Difference Learning in Humanoid Locomotion, IEEE RA-L 2025

Machine Learning

• Training Non-differentiable Networks via Optimal Transport, under review at TMLR
• AAC: Admissible-by-Architecture Differentiable Landmark Compression for ALT, under review at TMLR

Numerical Physics (GR)

• On the rarity of rocket-driven Penrose extraction in Kerr spacetime, Physical Review D 2026
• Observer-robust energy condition verification for warp drive spacetimes, under review at CQG

Interests & keywords

Vectorized robot learning & planning · Tensor search / batched optimization · OT & gradient flows · Diffusion / flow matching for motion generation · Humanoid locomotion & loco-manipulation · VLA / VLM for grasping & manipulation · Reproducible simulation + numerical GR

Links

• 🌐 LinkedIn: @anindex
• 🐦 Twitter/X: @an_thai_le
• 🎓 Google Scholar: An Thai Le
• 💌 Email: an@robot-learning.de