Values

The ad says“Strong grasp of computational Bayesian methods, including MCMC sampling methods and variational inference”
The ad says“Familiarity with deep learning theory and literature, including theory of over-parameterization and scaling laws”
The ad says“Experience training energy-based models (EBMs) or diffusion models”
The ad says“Experience training energy-based models (EBMs) or diffusion models”
The ad says“Experience with graph neural networks (GNNs) or graph message passing algorithms”
The ad says“Experience with graph neural networks (GNNs) or graph message passing algorithms”
The ad says“Strong theoretical background in information geometry”
The ad says“Experience with infrastructure for deep learning experimentation and training (Slurm, Ray, Kubernetes, Weights & Biases, etc.)”
The ad says“Strong grasp of computational Bayesian methods, including MCMC sampling methods and variational inference”
The ad says“Familiarity with deep learning theory and literature, including theory of over-parameterization and scaling laws”
The ad says“Experience with infrastructure for deep learning experimentation and training (Slurm, Ray, Kubernetes, Weights & Biases, etc.)”
The ad says“Familiarity with deep learning theory and literature, including theory of over-parameterization and scaling laws”
The ad says“Experience with infrastructure for deep learning experimentation and training (Slurm, Ray, Kubernetes, Weights & Biases, etc.)”
The ad says“Strong grasp of computational Bayesian methods, including MCMC sampling methods and variational inference”
The ad says“Experience with infrastructure for deep learning experimentation and training (Slurm, Ray, Kubernetes, Weights & Biases, etc.)”

Overview

Extropic is looking for junior ML scientists to join our residency program on either a part-time or full-time basis. This is a flexible program that can be similar to an internship (minimum 3 months) but we give our residents far more autonomy than most internships.

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 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

Preferred Qualifications

Familiarity with deep learning theory and literature, including theory of over-parameterization and scaling laws
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.)

Extropic is an equal opportunity employer

This position will require access to information subject to control under U.S. export control laws and regulations, including the Export Administration Regulations (“EAR”). Please note that any offer for employment will be conditioned on authorization to receive controlled items.

Asks for

Values