Manuel Vargas Guzmán

Manuel Vargas Guzmán

Logic and Deep Learning

Posts by Collection

publications

Model-Checking Process Equivalences

Published in Games, Automata, Logic and Formal Verification (GandALF), 2012

The paper introduces a modal fixpoint logic framework that uniformly defines many process equivalences (like bisimulation and trace equivalence). It enables symbolic model checking for a large fragment of the logic, and uses partial evaluation to derive decision procedures for verifying process equivalence.

Recommended citation: Martin Lange, Etienne Lozes, and Manuel Vargas Guzmán. (2012). "Model-Checking Process Equivalences." Games, Automata, Logic and Formal Verification (GandALF). Pages 43-56.
Download Paper

Model-checking process equivalences

Published in Theoretical Computer Science, 2014

This paper introduces a modal fixpoint logic framework that uniformly defines many process equivalences (like bisimulation and trace equivalence) and supports model checking in PTIME and PSPACE.

Recommended citation: Martin Lange, Etienne Lozes, and Manuel Vargas Guzmán. (2014). "Model-checking process equivalences." Theoretical Computer Science. 560(3). Pages 326-347.
Download Paper

Prepositions Matter in Quantifier Scope Disambiguation

Published in 29th International Conference on Computational Linguistics, 2022

This paper shows that incorporating prepositional sense information significantly improves machine learning models for quantifier scope disambiguation (QSD).

Recommended citation: Aleksander Leczkowski, Justyna Grudzińska, Manuel Vargas Guzmán, Aleksander Wawer, and Aleksandra Siemieniuk. (2022). "Prepositions Matter in Quantifier Scope Disambiguation." 29th International Conference on Computational Linguistics. Pages 3960–3970.
Download Paper

Compositionality in a simple corpus

Published in Journées Jointes des Groupements de Recherche Linguistique Informatique, Formelle et de Terrain (LIFT) et Traitement Automatique des Langues (TAL), 2022

We study whether neural networks can learn compositionality by training them on a simple propositional logic corpus

Recommended citation: Manuel Vargas Guzmán, Maria Boritchev, Jakub Szymanik, and Maciej Malicki. (2022). "Compositionality in a simple corpus." Journées Jointes des Groupements de Recherche Linguistique Informatique, Formelle et de Terrain (LIFT) et Traitement Automatique des Langues (TAL). Pages 55-63.
Download Paper

Testing the limits of logical reasoning in neural and hybrid models

Published in Annual Conference of the North American Chapter of the Association for Computational Linguistics, 2024

We evaluate how well neural models generalize syllogistic logical reasoning, using custom tests focused on compositionality and recursion.

Recommended citation: Manuel Vargas Guzmán, Jakub Szymanik, and Maciej Malicki (2024). "Testing the limits of logical reasoning in neural and hybrid models." Findings of the Association for Computational Linguistics: NAACL. Pages 2267–2279.
Download Paper

talks

teaching

Part-Time Lecturer – Universidad del Valle

Undergraduate courses, Universidad del Valle, Ingeniería de Sistemas Informáticos, 2020

Delivered undergraduate lectures from February 2018 to June 2020 in the Department of Computer Science Engineering at Universidad del Valle, La Paz, Bolivia.

Lecturer (Teaching Intern) - University of Warsaw

Undergraduate course, University of Warsaw, Faculty of Philosophy, 2025

Delivered lectures for first-cycle studies during the summer semesters (2021/22–2024/25) as part of the Cognitive Science program, coordinated by the Faculty of Philosophy, University of Warsaw.