At Symbolica, we are building deep learning models which perform structured reasoning: manipulate structured data, learn algebraic structure in it, and do so with an interpretable and verifiable logic. To that end, we are developing new mathematical foundations for deep learning: categorical deep learning. We are now assembling a R&D lab of expert category theory and machine learning researchers to develop this theory and apply it to the problems of code synthesis and theorem proving. We are committed to fundamental ideas, but also their execution in practice.

As a machine learning scientist, you will help us expand and carry out our research & development program. You will work in a team on projects developing categorical deep learning and experiment with novel architectures, helping us form and validate hypotheses. This is a rare opportunity to work on an innovative and transformative project, and make significant contributions to the field of artificial intelligence and applied category theory.

Responsibilities:

• Help the development of pioneering research projects in categorical deep learning
• Work closely with category theorists and machine learning researchers, bridging the gap between state of the art research on deep learning architectures and their structural formulation in category theory
• Stay at the forefront of experimental advances in deep learning, and ensure our theoretical models are grounded and coherent with respect to these advances
• Help us form hypotheses, and validate them experimentally by training and evaluating deep learning models
• Work simultaneously at different levels of abstraction - from understanding high-level categorical constructions to implementing low-level details of architecture in code

Preferred qualifications:

• PhD in Computer Science, Mathematics, or similar discipline.
• Proven track record of research published in top-tier conferences (e.g. NeurIPS, ICML, ICLR, AAAI, COLT), and journals
• Experience with functional programming languages (e.g. Haskell, Idris, Scala)
• Practical experience with deep learning frameworks such as JAX, Pytorch, Tensorflow, etc.
• Expertise in neural network architectures and a strong interest in category theory or type theory
• Exceptional communication and interpersonal skills.

Location:

• Melbourne (preferred) or AUS remote

We offer competitive compensation, including equity and health insurance. Salary and equity levels are commensurate with experience and location.