Algorithm Research Intern

Responsibilities

1. Under guidance, investigate generative models for Physical AI, including Diffusion Models / Diffusion Policies, encoder-decoder architectures, VLA/VLM, reinforcement learning, and world models. Analyze limitations and opportunities for improvement, and help propose new architectures that integrate quantum computing.
2. Help abstract specific problems into standard models or mathematical formulations suitable for quantum computing, such as Ising or QUBO. Conduct theoretical analysis, develop prototypes, and assist with solving and validating them on quantum computers.
3. Contribute to research on data refinement, augmentation, and quality evaluation, developing reusable methods and modules.
4. Help establish classical algorithm baselines and evaluation methods, validating the benefits and applicability limits of quantum approaches through reproducible controlled experiments.
5. Contribute to papers, technical reports, and other research outputs, and turn effective methods into reusable algorithm capabilities.
6. Collaborate with research and engineering teams, maintaining experiment records and reproducible results.

Requirements

1. Currently pursuing a PhD, or an outstanding research-focused master’s degree, in computer science, artificial intelligence, mathematics, physics, robotics, or a related field. At least six consecutive months of internship availability is mandatory; short internships are not offered. Availability on site for at least four days per week is preferred.
2. Strong programming skills: proficiency in Python and familiarity with deep learning frameworks such as PyTorch, with the ability to independently train models and implement experiments.
3. Strong mathematical modeling skills and the ability to abstract real-world problems into standard mathematical formulations. Familiarity with Ising or QUBO is preferred.
4. Proficiency with AI agent tools such as Codex or Claude Code, or experience building and using your own agents. You are welcome to demonstrate a project, experiment, or demo created in collaboration with AI during the interview.
5. Substantial experience in at least one of the following areas (one is sufficient; experience in all areas is not required):
   • Generative models: familiarity with at least one of Diffusion Models / Diffusion Policies, encoder-decoder architectures, VLA/VLM, reinforcement learning, or world models; an understanding of their fundamentals; and hands-on model training experience.
   • Robotics: research in high-fidelity simulation environments or validation of models on real robots. Experience in bimanual manipulation, dexterous hand manipulation, or whole-body motion control is preferred.
   • Quantum computing: academic or practical experience in quantum computing, QUBO/Ising modeling, or quantum-inspired methods.
6. The ability to independently read papers, formulate questions, design experiments, and implement code prototypes, with attention to experiment records and reproducibility.
7. Enthusiasm for the intersection of quantum computing and AI, and a willingness to continue learning quantum computing through interdisciplinary collaboration.

Preferred qualifications

• Publications at relevant conferences or journals, such as NeurIPS, ICML, ICLR, CoRL, RSS, ICRA, CVPR, ICCV, SIGGRAPH, T-RO, IJRR, or TPAMI; open-source research projects; or demonstrable algorithm innovations.
• Research or course project experience in energy-based models, sampling optimization, or related topics.

How to apply