Welcome to my little corner of the web! I’m Caleb, a computer scientist and logician currently completing my PhD at Indiana University. I hope to better understand reasoning and learning from a formal, mathematical perspective. This means that I work with formal models from logic, theoretical computer science, and cognitive science, and try to connect them to things such as: neural network learning, neuro-symbolic AI, belief revision, and human learning & development.
I also enjoy using dependently typed languages such as Lean & Agda, both as programming languages and proof assistants. I find that they help me organize, clarify, and focus my thoughts. I’m excited for the recent proof assistant tide in mathematics!
A Neural Network Model Builder, Verified in Lean A program that will build a “correct by construction” neural network that obeys constraints on its inference & learning. about | github
Learning Theory and Dynamic Epistemic Logic course A summer course (co-taught by Nina Gierasimczuk) where we try to unify the perspectives of computational learning theory, belief revision, dynamic epistemic logics, and neural network learning. – NASSLLI 2025:course link
Neural Network Semantics (Dissertation Topic) An approach to neuro-symbolic AI where logic operators are interpreted directly on a neural network. – Dissertation: draft (coming soon!) – FLaNN 2026:poster – AAAI 2024:paper | talk | poster | github – FLAIRS 2022:paper | talk | github
Grad School Era (old)
Notakto Player (Qualifying Exam Project) report | github
Neuro-Symbolic AI using Case-Based Reasoning – Class project:report | github – Class project:report | jupyter notebook – ICCBR 2019:paper | github
Syllogistic Logic and Natural Language Inference – Class project:report | github – AAAI 2020:paper | talk | github