Fourth Eon is developing adaptive, AI-native safeguards across the biotechnology stack, with a focus on function-based DNA synthesis screening. Fellows in this stream will work on technical research projects at the intersection of AI and biosecurity. Projects span topics like mechanistic interpretability of protein foundation models, bio model evaluations for biosecurity-relevant capabilities, and agentic sequence analysis workflows.
Fourth Eon is researching and developing adaptive, AI-native biological threat detection to improve the robustness of biosecurity safeguards throughout the biotechnology stack.
A core focus of our work is function-based screening of DNA synthesis. The following are a few examples of research projects that align with our research program, though I am open to exploring suggestions for project ideas from the applicants:
• Applying mechanistic interpretability techniques to protein foundation models to probe their learned representations of functionally meaningful biophysical and biochemical properties (e.g. local structure, surface charge, catalytic activity, solvation). The goal is to understand how semantic information is encoded in specific features and circuits for different models, and how to leverage them for reliable hazardous function prediction.
• Evaluating biological AI models for biosecurity-relevant capabilities, such as design of novel functional proteins, or biomolecular property prediction on non-natural genes and proteins.
• Developing agentic sequence analysis workflows for adaptive biosecurity screening. This project explores fine-tuning of AI agents for orchestrating computational biology tools and synthesizing the outputs to produce biosecurity risk assessments on unknown sequences.
As part of our Safe and Responsible Research Framework we require fellows to sign a fellowship agreement, which is a short, plain-language document intended to manage dual-use risks and information hazards related to sensitive biosecurity research. Fellows remain free to talk publicly about the nature of the fellowship and their project. Happy to discuss further or share the draft with the MATS team.
I typically schedule a standing weekly 1:1 meeting with each fellow, and also hold a weekly research group meeting. Beyond that I am available on Slack and can find additional time for calls outside of scheduled meetings.
Note that as part of our Safe and Responsible Research Framework we require fellows to sign a fellowship agreement covering confidentiality and pre-publication review for dual-use risks. This is common practice in biosecurity research and allows us to work freely together on sensitive material.
Required:
• Prior technical research experience
• Strong critical thinking and creative problem-solving abilities
• The integrity and judgment to responsibly carry out sensitive research
• A good understanding of the basics of biomolecular sequence, structure, and function
• Expertise in one or more of the following domains:
bioinformatics, computational biology, structural biology, biochemistry, molecular biophysics, protein engineering, biosecurity, AI/ML, or a related field
• Proficiency with Python
Preferred:
• Hands-on experience with testing biological AI models
• Have built model evaluations / benchmarks
• Experience with mechanistic interpretability techniques
• Biosecurity context awareness
Fellows who are interested in our research area should think of potential project ideas that leverage their strengths and interests. I will work individual fellows to identify a specific project that matches their background and interests and is aligned with our overall research direction, and to refine the scope and objectives of the project.