Autonomous AI Research Intern: World Models & VLA

Atoms; San Francisco (CA); On-site; USD 80,000 - 113,000; Full time

Job summary

Atoms seeks an AI/ML Research Intern to bridge high-level AI research with real-world autonomous transport platforms. You will contribute to core research on World Models and Vision-Language-Action models, helping translate breakthroughs into scalable deployment for perception, simulation, and navigation.

You will prototype end-to-end multimodal models, develop world models from sensor data, and ensure models adapt to urban edge cases. Mountain View or SF offices; onsite five days a week.

Qualifications

• PhD candidate in a quantitative field with strong research background.
• Publications or submissions in top ML/robotics venues is a plus.
• Proven ability to translate theory to real-world systems.

Responsibilities

• Research and develop RL and distillation methods for trajectory planning.
• Prototype end-to-end multimodal models translating perception to action.
• Build world models from multi-sensor logs for re-simulation.
• Collaborate with validation to test models in simulated environments.

Skills

PyTorch

JAX

Python

RL

Education

PhD in CS/Robotics

Tools

DeepSpeed

Megatron-LM

Ray

Job description

Atoms seeks an AI/ML Research Intern to bridge high-level AI research with real-world autonomous transport platforms. You will contribute to core research on World Models and Vision-Language-Action models, helping translate breakthroughs into scalable deployment for perception, simulation, and navigation.

You will prototype end-to-end multimodal models, develop world models from sensor data, and ensure models adapt to urban edge cases. Mountain View or SF offices; onsite five days a week.