Natural Language to First-Order Logic LLM-based Autoformalization
This paper addresses this gap: we first provide a principled definition for the FOL-autoformalization task by distinguishing Ontology Extraction from Logical Translation, showing how their conflation obscures (cross-study) evaluation; we review existing datasets, evaluation metrics, and LLM-based methods, including fine-tuning, prompting, and verification-based refinement; we identify open challenges in benchmarking, semantic evaluation, ontology-aware methods, and end-to-end applications.
Key points
- Large Language Models (LLMs) have renewed interest in autoformalization.
- Yet, when First-Order Logic (FOL) is considered as the target formalism, the field still lacks a unified task formulation and a systematic survey.
Sources (1)
- [1]Natural Language to First-Order Logic LLM-based AutoformalizationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 02:24 PM
This paper addresses this gap: we first provide a principled definition for the FOL-autoformalization task by distinguishing Ontology Extraction from Logical Translation, showing how their conflation obscures (cross-study) evaluation; we review existing datasets, evaluation metrics, and LLM-based methods, including fine-tuning, prompting, and verification-based refinement; we identify open challen
Large Language Models (LLMs) have renewed interest in autoformalization.
Extractive summary: sentences quoted from the sources.