Member of the Technical Staff, Interpretability

Output Biosciences•New York, NY

4d•$120,000 - $250,000

About The Position

Output has built a biological reasoning model that understands biology at the scale and complexity life actually operates. Our model independently learned the principles of molecular interactions, opening up drug treatments that were previously impossible. We're already generating therapies that traditional approaches cannot reach. The hardest problems in both AI and biology are being solved here, and there is room for you to own one. Output is currently in stealth, operated by a team of repeat founders and biotech veterans with multiple exits in AI x Bio, and backed by top-tier VCs including Y Combinator. You will continue developing methods to understand what our foundation model learns about biology, and build the tools that make it a glass box model. We believe that in biology, a model's reasoning must be visible. And the features you find are not just explanations: they expand what the model can do.

Requirements

PhD in computer science, machine learning, physics, mathematics, or a related field with 2+ years of post-doctoral or industry research experience, OR a Bachelor's or Master's degree with 5+ years of hands-on research and engineering experience in model interpretability or representation analysis
Strong publication record at top-tier venues (e.g., NeurIPS, ICML, ICLR) with contributions to mechanistic interpretability, representation analysis, probing methods, or model understanding
Hands-on experience analyzing the internal representations of large neural networks, with demonstrated ability to design experiments that reveal what models have learned
Proficient in Python and PyTorch, and have experience working with large models on GPU infrastructure
Demonstrated ability to take interpretability research from experiments to usable tools: you do not just analyze models, you build systems others can use
Write production-quality code that is well-tested and maintainable, and you are comfortable working in shared codebases with version control and code review
Think carefully about what constitutes evidence that a model has learned a concept, and you design experiments that distinguish real capabilities from artifacts

Nice To Haves

Background in chemistry, biology, computational biology, biophysics, or a related natural science
Experience interpreting ML models trained on scientific or biological data
Experience building visualization or analysis tools for model internals
Experience with multimodal models or representations that span multiple data types
Contributed to open-source machine learning projects

Responsibilities

Continue developing our methods for probing and reverse-engineering the model's learned representations, understanding how it encodes biological information across molecular scales
Design and run experiments to identify and characterize capabilities, mapping what the model has learned about molecular interactions and biological function
Build methods to extract the model's biological understanding as explicit, usable outputs that downstream systems and researchers can act on
Create tools that connect model internals to meaningful biological concepts, making the model's reasoning interpretable to scientists and useful in practice
Work closely with the pretraining and generation teams, feeding interpretability findings back into model development to strengthen the capabilities you uncover
Own the full pipeline from probing experiments to production-quality interpretability tools, building robust systems on distributed infrastructure