Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges

State-space graphs and automata serve as fundamental tools for modeling and analyzing the behavior of computational systems. Recurrent neural networks (RNNs) and language models are deeply intertwined, as RNNS provide the foundational architecture that enables language models to process sequential d...

Full description

Saved in:
Bibliographic Details
Main Authors: Chanon Dechsupa, Teerapong Panboonyuen, Wiwat Vatanawood, Praisan Padungweang, Chakchai So-In
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10993355/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849715219336527872
author Chanon Dechsupa
Teerapong Panboonyuen
Wiwat Vatanawood
Praisan Padungweang
Chakchai So-In
author_facet Chanon Dechsupa
Teerapong Panboonyuen
Wiwat Vatanawood
Praisan Padungweang
Chakchai So-In
author_sort Chanon Dechsupa
collection DOAJ
description State-space graphs and automata serve as fundamental tools for modeling and analyzing the behavior of computational systems. Recurrent neural networks (RNNs) and language models are deeply intertwined, as RNNS provide the foundational architecture that enables language models to process sequential data, capture contextual dependencies, and improve natural language processing tasks. Both RNNs and state-space graphs help evaluate discrete-time systems within this formal framework. However, the basic question of their equivalence remains an open challenge regarding the models governing sentence structure in natural language and the formal model in automata theory. In this paper, we present ENGRU (Enhanced Gated Recurrent Units), a novel deep learning-based approach for formal verification. ENGRU integrates concepts of model checking techniques, Colored Petri Nets (CPNs), and sequential learning to analyze systems at a high level of abstraction. CPN models undergo state-space enumeration via a model-checking tool, generating a state-space graph and an automaton from inherent state transition patterns. These graphs are transformed into sequential representations as sub-paths, enabling ENGRU to learn the execution paths and predict system behaviors. ENGRU effectively predicts goal states within discrete-time models by the competency of gated recurrent mechanisms, encouraging early bug detection and allowing predictive state-space exploration. Experimental results demonstrate that ENGRU&#x2019;s ability achieves high accuracy and efficiency in goal state predictions. The source code for ENGRU is available at (<uri>https://github.com/kaopanboonyuen/ENGRU</uri>).
format Article
id doaj-art-ea0cf6b3e5294367b44c9a4b0ec902f8
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-ea0cf6b3e5294367b44c9a4b0ec902f82025-08-20T03:13:29ZengIEEEIEEE Access2169-35362025-01-0113843578437910.1109/ACCESS.2025.356819410993355Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its ChallengesChanon Dechsupa0https://orcid.org/0000-0001-8441-8962Teerapong Panboonyuen1https://orcid.org/0000-0001-8464-4476Wiwat Vatanawood2https://orcid.org/0000-0003-0457-3474Praisan Padungweang3https://orcid.org/0000-0002-4358-7927Chakchai So-In4https://orcid.org/0000-0003-1026-191XDepartment of Computer Science, College of Computing, Khon Kaen University, Khon Kaen, ThailandDepartment of Computer Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, ThailandDepartment of Computer Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, ThailandDepartment of Computer Science, College of Computing, Khon Kaen University, Khon Kaen, ThailandDepartment of Computer Science, College of Computing, Khon Kaen University, Khon Kaen, ThailandState-space graphs and automata serve as fundamental tools for modeling and analyzing the behavior of computational systems. Recurrent neural networks (RNNs) and language models are deeply intertwined, as RNNS provide the foundational architecture that enables language models to process sequential data, capture contextual dependencies, and improve natural language processing tasks. Both RNNs and state-space graphs help evaluate discrete-time systems within this formal framework. However, the basic question of their equivalence remains an open challenge regarding the models governing sentence structure in natural language and the formal model in automata theory. In this paper, we present ENGRU (Enhanced Gated Recurrent Units), a novel deep learning-based approach for formal verification. ENGRU integrates concepts of model checking techniques, Colored Petri Nets (CPNs), and sequential learning to analyze systems at a high level of abstraction. CPN models undergo state-space enumeration via a model-checking tool, generating a state-space graph and an automaton from inherent state transition patterns. These graphs are transformed into sequential representations as sub-paths, enabling ENGRU to learn the execution paths and predict system behaviors. ENGRU effectively predicts goal states within discrete-time models by the competency of gated recurrent mechanisms, encouraging early bug detection and allowing predictive state-space exploration. Experimental results demonstrate that ENGRU&#x2019;s ability achieves high accuracy and efficiency in goal state predictions. The source code for ENGRU is available at (<uri>https://github.com/kaopanboonyuen/ENGRU</uri>).https://ieeexplore.ieee.org/document/10993355/Model checkingcolored petri netstate-space analysisGRU attentionformal verification
spellingShingle Chanon Dechsupa
Teerapong Panboonyuen
Wiwat Vatanawood
Praisan Padungweang
Chakchai So-In
Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
IEEE Access
Model checking
colored petri net
state-space analysis
GRU attention
formal verification
title Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
title_full Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
title_fullStr Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
title_full_unstemmed Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
title_short Toward AI-Augmented Formal Verification: A Preliminary Investigation of ENGRU and Its Challenges
title_sort toward ai augmented formal verification a preliminary investigation of engru and its challenges
topic Model checking
colored petri net
state-space analysis
GRU attention
formal verification
url https://ieeexplore.ieee.org/document/10993355/
work_keys_str_mv AT chanondechsupa towardaiaugmentedformalverificationapreliminaryinvestigationofengruanditschallenges
AT teerapongpanboonyuen towardaiaugmentedformalverificationapreliminaryinvestigationofengruanditschallenges
AT wiwatvatanawood towardaiaugmentedformalverificationapreliminaryinvestigationofengruanditschallenges
AT praisanpadungweang towardaiaugmentedformalverificationapreliminaryinvestigationofengruanditschallenges
AT chakchaisoin towardaiaugmentedformalverificationapreliminaryinvestigationofengruanditschallenges