Herein is code for a vector-symbolic implementation of a Lisp interpreter. The interpreter is admittedly slow and inefficient and the Lisp is restricted to the minimal operators sufficient for Turing-completeness. But the code constitutes a proof-of-concept that Lisp can be implemented using a vector-symbolic architecture.

The interpreter is described formally in the following paper. If you are looking to cite this code, cite this paper:

Flanagan, E.T. & Kelly, M. A. (2024). Hey Pentti, we did it!: A fully vector-symbolic Lisp. In Proceedings of the 22nd International Confernece on Cognitive Modeling. Tilburg, the Netherlands: Tilburg University.

The code is written by E. Tomkins-Flanagan and M. A. Kelly by modifying the minimal Lisp interpreter written in Python by Peter Norvig.

The original code by Norvig can be found here:

https://www.norvig.com/lispy.html

It's a Jupyter notebook. Open it in Jupyter. That's it.

You can also upload it to Google Colab.

Trust me, you don't want this.

A copy can be found in this repository, named heypentti.pdf.