BAIT: Benchmarking (Embedding) Architectures for Interactive Theorem-Proving

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lamont, Sean, Norrish, Michael, Dezfouli, Amir, Walder, Christian, Montague, Paul
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913688890376192
author Lamont, Sean
Norrish, Michael
Dezfouli, Amir
Walder, Christian
Montague, Paul
author_facet Lamont, Sean
Norrish, Michael
Dezfouli, Amir
Walder, Christian
Montague, Paul
contents Artificial Intelligence for Theorem Proving has given rise to a plethora of benchmarks and methodologies, particularly in Interactive Theorem Proving (ITP). Research in the area is fragmented, with a diverse set of approaches being spread across several ITP systems. This presents a significant challenge to the comparison of methods, which are often complex and difficult to replicate. Addressing this, we present BAIT, a framework for fair and streamlined comparison of learning approaches in ITP. We demonstrate BAIT's capabilities with an in-depth comparison, across several ITP benchmarks, of state-of-the-art architectures applied to the problem of formula embedding. We find that Structure Aware Transformers perform particularly well, improving on techniques associated with the original problem sets. BAIT also allows us to assess the end-to-end proving performance of systems built on interactive environments. This unified perspective reveals a novel end-to-end system that improves on prior work. We also provide a qualitative analysis, illustrating that improved performance is associated with more semantically-aware embeddings. By streamlining the implementation and comparison of Machine Learning algorithms in the ITP context, we anticipate BAIT will be a springboard for future research.
format Preprint
id arxiv_https___arxiv_org_abs_2403_03401
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BAIT: Benchmarking (Embedding) Architectures for Interactive Theorem-Proving
Lamont, Sean
Norrish, Michael
Dezfouli, Amir
Walder, Christian
Montague, Paul
Artificial Intelligence
Machine Learning
Logic in Computer Science
Artificial Intelligence for Theorem Proving has given rise to a plethora of benchmarks and methodologies, particularly in Interactive Theorem Proving (ITP). Research in the area is fragmented, with a diverse set of approaches being spread across several ITP systems. This presents a significant challenge to the comparison of methods, which are often complex and difficult to replicate. Addressing this, we present BAIT, a framework for fair and streamlined comparison of learning approaches in ITP. We demonstrate BAIT's capabilities with an in-depth comparison, across several ITP benchmarks, of state-of-the-art architectures applied to the problem of formula embedding. We find that Structure Aware Transformers perform particularly well, improving on techniques associated with the original problem sets. BAIT also allows us to assess the end-to-end proving performance of systems built on interactive environments. This unified perspective reveals a novel end-to-end system that improves on prior work. We also provide a qualitative analysis, illustrating that improved performance is associated with more semantically-aware embeddings. By streamlining the implementation and comparison of Machine Learning algorithms in the ITP context, we anticipate BAIT will be a springboard for future research.
title BAIT: Benchmarking (Embedding) Architectures for Interactive Theorem-Proving
topic Artificial Intelligence
Machine Learning
Logic in Computer Science
url https://arxiv.org/abs/2403.03401