Research Scientist Intern, 3D Vision & World Simulation (PhD)

About the Role

Reality Labs Research is seeking a Research Scientist Intern to help develop the next generation of intelligent assistance systems that guide users in contextual and adaptive environments.

We are looking for candidates with expertise in AI-driven 3D spatial understanding and generative modeling, including experience building and iterating on deep learning systems at the intersection of:

• Generative AI (e.g., diffusion models)

• World simulation and modeling

• 3D computer vision and reconstruction

Ideal candidates have experience with predicting physics and future states, camera geometry, depth estimation, and motion or trajectory forecasting.

Responsibilities

• Develop, implement, and evaluate methods to improve the performance and interpretability of Vision-Language Models (VLMs) and related AI/ML models.

• Write modular, reusable research code and leverage Meta's large-scale infrastructure to accelerate experimentation.

• Collaborate with researchers and engineers to prototype, train, and evaluate models at scale.

• Deliver creative, well-engineered solutions to challenging research problems.

• Produce research suitable for publication at leading conferences and journals, including:

  • NeurIPS

  • ICLR

  • CVPR

  • ECCV

  • ICML

  • ICCV

  • AAAI

  • IJCAI

  • ICRA

  • IEEE TPAMI

  • IJCV

  • IEEE RA-L

• Design and develop generative AI models (e.g., diffusion models) to:

  • Learn environment dynamics

  • Predict future states

  • Simulate physical interactions within world models

• Build and advance 3D computer vision and reconstruction pipelines using:

  • Camera geometry

  • Depth estimation

  • Related geometric techniques

• Research and implement methods for 3D spatial understanding, including:

  • Motion forecasting

  • Trajectory prediction

  • Spatial reasoning about objects and agents

Minimum Qualifications

• Strong programming skills in Python.

• Extensive hands-on experience building, training, and debugging deep learning models using PyTorch.

• Experience with generative AI architectures, including diffusion models.

• Understanding of world modeling concepts, including:

  • Learning environment dynamics

  • Predicting future states

  • Simulating physical interactions

• Strong foundation in 3D computer vision and 3D reconstruction, including:

  • Camera geometry

  • Depth estimation

• Experience working with:

  • 3D spatial data

  • Motion forecasting

  • Trajectory prediction

Preferred Qualifications

• Experience with advanced generative modeling approaches, including:

  • Flow Matching

  • Diffusion Transformers (DiTs)

• Hands-on experience with modern 3D representations, including:

  • 3D Gaussian Splatting (3DGS)

  • Dense 3D point trajectory forecasting in world coordinates

• Background in robotics planning and manipulation, including transferring learned representations to downstream tasks such as:

  • Closed-loop pick-and-place

  • Real-world robotic control

• Familiarity with Vision-Language Models (VLMs), including:

  • Object grounding

  • Natural language instruction following

  • Multi-modal token processing

• Experience extending world models to physical applications, including:

  • Latent-action world models

  • Trajectory-conditioned video generation

• Strong publication record at leading conferences such as:

  • CVPR

  • ICCV

  • ECCV

  • ICLR

  • NeurIPS

  • ICRA

  • RSS

Compensation

$7,650–$12,134 per month + benefits

Individual compensation is determined by skills, qualifications, experience, and work location. The compensation listed reflects base pay only and does not include bonuses, equity, or other incentives. Eligible interns also receive Meta's comprehensive benefits package.

About Meta

Meta builds technologies that help people connect, find communities, and grow businesses. Since the launch of Facebook in 2004, Meta has expanded its family of products to include Messenger, Instagram, and WhatsApp, connecting billions of people worldwide.

Today, Meta is focused on building the next generation of computing through augmented and virtual reality, creating immersive experiences that move beyond traditional 2D screens and redefine how people interact with digital and physical worlds.

Equal Opportunity

Meta is an Equal Employment Opportunity and Affirmative Action employer. Employment decisions are made without regard to race, religion, color, national origin, sex (including pregnancy and related medical conditions), sexual orientation, gender identity or expression, age, veteran status, disability, or any other legally protected characteristic.

Qualified applicants with criminal histories will also be considered in accordance with applicable federal, state, and local laws. Meta participates in the E-Verify program where required.

Meta may leverage artificial intelligence and machine learning technologies during the hiring process.

Accommodations

Meta is committed to providing reasonable accommodations throughout the recruiting process. Candidates requiring accommodations due to a disability may contact:

accommodations-ext@meta.com