Thermo ML Resident
San Francisco
ML Team /
Hybrid
Overview
Extropic is looking for junior ML scientists to join our residency program on either a part-time or full-time basis. Our hardware massively accelerates certain kinds of probabilistic inference, and residents will help pioneer the science of training models in the thermodynamic paradigm.
Responsibilities
- Collaborate with senior researchers to derive the theory of new probabilistic models and their learning rules, including energy-based models and diffusion models
- Scale up experimentation infrastructure and optimize over the design space of models
- Implement, visualize, and evaluate new architectures, training algorithms, and benchmarks
- Publish papers, contribute to open source, and communicate design insights to our hardware team
Required Qualifications
- Experience in scientific Python
- Experience with JAX or similar deep learning framework (PyTorch, TensorFlow, or Keras)
- Strong foundations in probability and linear algebra
- Projects or papers demonstrating hands-on experience in applied machine learning and data science
- Familiarity with deep learning theory and literature, including theory of over-parameterization and scaling laws
Preferred Qualifications
- Experience training energy-based models (EBMs) or diffusion models
- Experience with graph neural networks (GNNs) or graph message passing algorithms
- Experience with infrastructure for deep learning experimentation and training (Slurm, Ray, Kubernetes, Weights & Biases, etc.)
- Strong theoretical background in information geometry
- Strong grasp of computational Bayesian methods, including MCMC sampling methods and variational inference
- Publications in top ML conferences (NeurIPS, ICML, ICLR, CVPR, etc.)
$75,000 - $150,000 a year
Salary and equity compensation will vary with experience
Extropic is an equal opportunity employer
