Synthesizing Neurosymbolic Programs
We discuss the emerging problem of automatically synthesizing neurosymbolic programs, or programs that blend together neural networks and high-level code.
Understanding the World with Program Synthesis
Most applications of program synthesis are concerned with the engineering of software. However, because programming languages can be used to model the physical world, program synthesis can also offer a way of discovering and validating new hypotheses in the natural sciences. In this post, I elaborate on how.
AI Safety as a PL problem
In the last few years, there have been many anxious ruminations about AI safety. To a significant extent, these fears come from the realization that modern AI systems can be alarmingly brittle. For example, a deep neural net used for image classification can produce...