Research Scientist, Program Synthesis & Neuro-symbolic Methods

About the Role

The role of Research Scientist, Program Synthesis & Neuro-symbolic Methods at Basis is centered on developing computational theories that explain how agents can generate and execute programs while reasoning about the physical world. You will advance program synthesis and neuro-symbolic methods to create systems that form and test scientific hypotheses in embodied settings. This involves designing mechanisms that integrate symbolic reasoning with neural representations to produce reliable and interpretable control. You will work at the intersection of theory and implementation, translating mathematical principles into working robotic behaviors. Collaboration with interdisciplinary teams is central, as you will define the fundamental limits of learning from interaction and drive projects from abstract ideas through to validated systems deployed on physical hardware.

What You'll Do

You will formulate research questions that connect program synthesis with neuro-symbolic reasoning in embodied intelligence. This includes architecting mechanisms for agents to learn causal models and interpretable programs through interaction with the physical world. You will engineer algorithms that bridge high-level symbolic reasoning with low-level robotic control and real-time execution. A core responsibility is implementing systems that generate verifiable control programs instead of opaque, black-box policies. You will conduct experiments that probe the limits of sample efficiency, generalization, and robustness in world modeling. You will partner with hardware engineers to validate theories on physical robot platforms and close the sim-to-real gap. You will translate findings from theoretical analysis into practical tools that accelerate scientific discovery and real-world deployment. Coordination with domain experts will help identify high-impact applications in manufacturing, science, and complex operational environments. Maintaining rigorous experimental standards that connect theoretical insights to measurable system behavior is essential, as is documenting and communicating methods so that results are reusable and extendable by interdisciplinary collaborators.

Requirements

To qualify for this role, you must have demonstrated an ability to conduct scientific research of high quality. Evidence of this can include publications at top venues such as NeurIPS, ICML, ICLR, POPL, and PLDI, as well as technical reports and impactful software projects. You must possess deep expertise in Program Synthesis & Neuro-symbolic Methods, including domain-specific languages, program induction, verifiable control, and neuro-symbolic integration. Strong mathematical and computational foundations are required, including probability theory, optimization, linear algebra, and the ability to implement complex algorithms from first principles. You must be comfortable working across the research-to-deployment pipeline, from theoretical development through experimental validation. Progress with autonomy and intellectual curiosity is expected, enabling you to identify valuable research directions within the broader MARA mission and drive projects to completion. Valuing collaboration and knowledge transfer, and actively sharing insights across specialization boundaries, is critical. You must be excited about solving real-world problems through embodied intelligence that advances society's ability to solve intractable problems. Commitment to a collaborative work style that thrives in a nonprofit research environment focused on building a new technological foundation guided by human values is required.

Nice to Have

A PhD (or equivalent experience) in technical areas including robotics, machine learning, computer vision, control theory, cognitive science, or physics is advantageous. Experience at leading robotics or AI labs, whether academic or industry, is beneficial. A track record of algorithms deployed on physical robot systems demonstrates practical impact. Contributions to major open-source projects in robotics or ML highlight community engagement. Experience with both theoretical research and systems engineering, and a background spanning multiple specialization areas within program synthesis and neuro-symbolic methods, are also advantageous.

Practical Notes

Location is restricted to the New York Office. Engagement is full-time. The role involves collaboration with both internal and external partners on problems larger than those manageable by a single person, reflecting Basis's commitment to a collaborative, values-driven research culture.