A Semantics for Belief in Simplicial Complexes

Published in Proceedings of AiML 2026, 2026

This is the first paper I have really published. In it, my advisor, Adam Bjorndahl and myself, set out to do two things. The first is to provide (yet another) semantics for belief in simplicial complexes. The internal motivation is that our approach is very simple, and preserves many existing, standard assumptions. However, the real motivation for this approach, as I see it, is that it also lets us do belief revision in the way I wanted for the joint work with NASA. The second thing this paper sets out to do is give a general technique for translating ``improper'' relational models into ``proper'' ones that preserves relevant canonical properties like transitivity and reflexivity. Doing this not only lets us prove completeness for our simplicial models, it likely closes a few other completeness conjectures in the existing literature.

Recommended citation: Adam Bjorndahl and Philip Sink (2026). "A Semantics for Belief in Simplicial Complexes" Proceedings of AiML 2026
Download Paper