LTLGuard: Formalizing LTL Specifications with Compact Language Models and Lightweight Symbolic Reasoning

TL;DR

LTLGuard is a resource-efficient toolchain that combines small language models with symbolic reasoning to accurately translate informal requirements into consistent LTL specifications.

Contribution

It introduces a modular approach integrating constrained generation and formal checking to improve LTL specification accuracy from natural language.

Findings

Effective generation of conflict-free LTL specifications

Improved accuracy over baseline language models

Demonstrated usability through quantitative evaluation

Abstract

Translating informal requirements into formal specifications is challenging due to the ambiguity and variability of natural language (NL). This challenge is particularly pronounced when relying on compact (small and medium) language models, which may lack robust knowledge of temporal logic and thus struggle to produce syntactically valid and consistent formal specifications. In this work, we focus on enabling resource-efficient open-weight models (4B--14B parameters) to generate correct linear temporal logic (LTL) specifications from informal requirements. We present LTLGuard, a modular toolchain that combines constrained generation with formal consistency checking to generate conflict-free LTL specifications from informal input. Our method integrates the generative capabilities of model languages with lightweight automated reasoning tools to iteratively refine candidate specifications, understand the origin of the conflicts and thus help in eliminating inconsistencies. We demonstrate the usability and the effectiveness of our approach and perform quantitative evaluation of the resulting framework.