Research Scientist, Program Synthesis & Neuro-symbolic Methods

About Basis

Basis is a nonprofit applied AI research organization with two mutually reinforcing goals.

The first is to understand and build intelligence. This means to establish the mathematical principles of what it means to reason, to learn, to make decisions, to understand, and to explain; and to construct software that implements these principles.

The second is to advance society’s ability to solve intractable problems
. This means expanding the scale, complexity, and breadth of problems that we can solve today, and even more importantly, accelerating our ability to solve problems in the future.

To achieve these goals, we’re building both a new technological foundation that draws inspiration from how humans reason, and a new kind of collaborative organization that puts human values first.

Research Scientists on the MARA (Modeling, Abstraction, and Reasoning Agents) project develop computational theories of scientific reasoning applied to robotics and embodied intelligence. You will advance the frontiers of world modeling, reinforcement learning, program synthesis, and robotic control to create systems that can learn, reason about, and interact with the physical world.

We are looking for exceptional researchers with expertise in Program Synthesis & Neuro-symbolic Methods
. The ideal candidate has a strong publication record in relevant venues, combines theoretical depth with practical implementation skills, and is excited about building systems that learn like scientists—forming hypotheses, conducting experiments, and building models of how the world works.

You will work as part of an interdisciplinary team tackling fundamental questions:
How can agents learn causal models from interaction? How do we bridge high-level reasoning with low-level control? How can we generate interpretable, verifiable control programs rather than black-box policies?

Basis is a collaborative effort, both internally and with our external partners; we are looking for people who enjoy working with others on problems larger than ones they can tackle alone.

• Have demonstrated an ability to do scientific research that is of high quality
  . Possible ways to demonstrate this include publications at top venues (NeurIPS, ICML, ICLR, POPL, PLDI), technical reports, and impactful software projects.

• Possess deep expertise in Program Synthesis & Neuro-symbolic Methods
  :

  • Domain-specific languages, program induction, verifiable control, neuro-symbolic integration.

  • Experience with combining neural networks with symbolic reasoning or program generation.

• Have strong mathematical and computational foundations including probability theory, optimization, linear algebra, and the ability to implement complex algorithms from first principles.

• Be comfortable working across the research-to-deployment pipeline
  , from theoretical development through experimental validation.

• Progress with autonomy and intellectual curiosity
  . You can identify valuable research directions within the broader MARA mission, design experiments, and drive projects to completion.

• Value collaboration and knowledge transfer
  . You actively share insights across specialization boundaries and help integrate diverse approaches into coherent systems.

• Be excited about solving real-world problems through embodied intelligence that advances our ability to understand and interact with the physical world.

In addition, the following would be an advantage:

• PhD (or equivalent experience) in technical areas including: robotics, machine learning, computer vision, control theory, cognitive science, or physics.

• Experience at leading robotics or AI labs (academic or industry).

• Track record of algorithms deployed on physical robot systems.

• Contributions to major open-source projects in robotics or ML.

• Experience with both theoretical research and systems engineering.

• Background spanning multiple specialization areas.

• Develop computational theories of intelligence specific to program synthesis and neuro-symbolic methods, focusing on synthesizing control programs, learning interpretable models, or bridging symbolic reasoning with neural learning.

• Design and implement…