Improving the Diproche CNL through Autoformalization via Large Language Models
This work addresses the challenge of improving automated proof-checking for educational purposes, though it appears incremental as it builds on an existing system with preliminary findings.
The paper tackled the problem of autoformalizing controlled natural language proofs in the Diproche system by using large language models, achieving encouraging initial results.
The Diproche system is an automated proof checker for texts written in a controlled fragment of German, designed for didactical applications in classes introducing students to proofs for the first time. The first version of the system used a controlled natural language for which a Prolog formalization routine was written. In this paper, we explore the possibility of prompting large language models for autoformalization in the context of Diproche, with encouraging first results.