Astera
Machine Learning Researcher - Springtail
About this role
Astera Institute seeks a Machine Learning Researcher to develop data-efficient architectures and novel approaches to model induction, including program synthesis and self-learning systems. This full-time role focuses on advancing ML's sample efficiency for small and expensive datasets common in scientific research.
What you'll do
- Design and validate improvements to architectural components like attention mechanisms through controlled experiments and dataset analysis
- Investigate runtime inference in gradient-trained networks using statistical learning theory and constrained optimization techniques
- Develop and maintain well-documented, performant code supporting experimental research
- Contribute to bootstrapped program synthesis and meta-learning system components
- Analyze learned representations and learning dynamics across different model configurations
What they're looking for
- Machine learning research and theory
- PyTorch proficiency
- JAX, CUDA, and/or Triton experience
- Mathematical modeling and statistical learning
- Experimental design and empirical validation
- Python and scientific computing
- Collaborative research and teamwork
Benefits
- Competitive compensation package based on Bay Area location
- Hybrid work arrangement in Emeryville, CA
- Opportunity to conduct fundamental ML research on ambitious problems
- Access to institutional resources and mission-driven environment
- Occasional travel for team collaboration
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Astera
Astera builds advanced infrastructure for large-scale distributed simulations, neuroscience research tools, and knowledge-sharing platforms aimed at accelerating scientific discovery. The company is hiring software engineers, ML researchers, and scientist-engineers to develop high-performance systems, specialized neuroscience instrumentation, and data-efficient machine learning approaches.
View all jobs at AsteraLikely interview questions
- Can you walk us through a recent research project where you hypothesized and tested improvements to a neural network architecture? What controlled experiments did you design?
- Describe your experience with PyTorch and how you've used it in research. Have you worked with JAX, CUDA, or Triton, and in what contexts?