- Le 11/12/2023 à 14h à St Charles, FRUMAM, 3ème étage
Orateur : Han Gao (LIS, UMR 7020, LIRICA)
Titre : Variants of intuitionistic modal logic
Résumé : Intuitionistic modal logic as a branch of non-classical logic enjoys a long history tracing back to 1940s. In the development since then, two main traditions have emerged, one called intuitionistic modal logic and the other constructive modal logic. These two traditions are motivated by meta-logical properties and applications in computer science respectively.
In this seminar, I will first give a brief introduction about the general development of the field of intuitionistic modal logic focusing on some prominent logics like IK, CCDL, CK, etc. Then I will introduce a recent work that reveals intuitionistic modal logics defined by natural variants of Kripke's semantics for which little or no research has been done. I will present the results we obtained so far concerning axiomatization, proof-systems and decidability of some of these logics. Finally, I will also discuss some open problems.
## Reference
P. Balbiani, H. Gao, Ç. Gencer and N. Olivetti, A
Natural Intuitionistic Modal Logic: Axiomatization and Bi-nestedCalculus, Computer Science Logic 2024 (CSL'24), to appear.
- Le 20/11/2023 à 14h à St Charles, FRUMAM, 3ème étage
Orateur : Van-Giang Trinh (LIS, UMR 7020, LIRICA)
Titre : Efficient enumeration of fixed points in complex Boolean networks using answer set programming
Résumé : Boolean Networks (BNs) are an efficient modeling formalism with applications in various research fields such as mathematics, computer science, and more recently systems biology. One crucial problem in the BN research is to enumerate all fixed points, which has been proven crucial in the analysis and control of biological systems. Indeed, in that field, BNs originated from the pioneering work of R. Thomas on gene regulation and from the start were characterized by their asymptotic behavior: complex attractors and fixed points. The former being notably more difficult to compute exactly, and specific to certain biological systems, the computation of stable states (fixed points) has been the standard way to analyze those BNs for years. However, with the increase in model size and complexity of Boolean update functions, the existing methods for this problem show their limitations. To our knowledge, all the state-of-the-art methods for the fixed point enumeration problem rely on Answer Set Programming (ASP). Motivated by these facts, in this work we propose two new efficient ASP-based methods to solve this problem. We evaluate them on both real-world and pseudo-random models, showing that they vastly outperform four state-of-the-art methods as well as can handle very large and complex models.
## Reference
Van-Giang Trinh, Belaid Benhamou, and Sylvain Soliman. "Efficient Enumeration of Fixed Points in Complex Boolean Networks Using Answer Set Programming." In 29th International Conference on Principles and Practice of Constraint Programming (CP 2023). Schloss Dagstuhl-Leibniz-Zentrum für Informatik, 2023.
- Le 18/09/2023 à 15h à St Charles, FRUMAM, 2ème étage
Orateur : Simon Vilmin (LIS, UMR 7020, LIRICA)
Titre : Enumération des modèles maximaux d'une formule de Horn
Résumé :L'énumération des modèles maximaux d'une formule de Horn est un problème difficile qui apparaît implicitement dans plusieurs domaines de l'informatique, allant du data mining à l'optimisation combinatoire.
En outre, il généralise la dualisation des fonctions monotones booléennes (ou dualisation des hypergraphes) et se rapporte au problème de dualisation dans les treillis codés par des bases d’implications. Dans cet exposé, je présenterai le problème et quelques résultats connus à son sujet.
En général, le problème ne peut être résolu en temps total polynomial à moins que P = NP, ce qui suggère l'étude de cas particuliers.Je m'intéresserai au cas où les clauses négatives de la formule ont taille deux.
En tirant parti du lien entre formules de Horn et treillis, j'utiliserai le nombre de Carathéodory et la dualisation des hypergraphes pour donner des classes de formules où le problème peut être résolu en temps total (quasi-)polynomial.
Je mentionnerai aussi des classes de treillis où le problème est déjà plus dur que la dualisation des hypergraphes, même si les clauses négatives sont de tailles deux.
- Le 10/07/2023 à 14h à St Charles, FRUMAM, 2ème étage
Orateurs : Vincent Risch (LIS, UMR 7020), Jean-Jacques Pinto (Aix-Marseille Université)
Titre : Une méthode en cours de formalisation : l'Analyse des Logiques Subjectives (A.L.S.)
Résumé :
– ANALYSE… de discours, explorant l'argumentation subjective (persuader plutôt que convaincre) en recherchant le dénominateur commun à l'auteur d'un texte et à ses auditeurs/lecteurs : qui l'apprécie, qui le rejette, et dans quels termes. Exemple traité au départ: diagnostiquer la sympathie-complicité permet de comprendre le lien subliminal entre "propagandeur" et "propagandé" et l’admissibilité des fake news.
– … des LOGIQUES… non pas "syntaxiques", mais "lexicales" : dans quels paradigmes le locuteur puise les éléments de son plaidoyer lexical. Alors que le raisonnement visant la conviction relève plutôt de la syntaxe, le type de persuasion analysé ici relève du lexique ("l'argumentation dans la langue" d'O. Ducrot), lexique porteur de valeurs en tant que "sténographie sémantique" d'inférences implicites..
– … SUBJECTIVES : complémentarité argumentative entre le régime cognitif (conscient et révisable) et le régime subjectif (non-conscient et non révisable). Le modèle de l'A.L.S. n'est pas binaire ("rationnel/irrationnel") mais ternaire : "rationnel/irrationnel/logique".
▪L’analyse rationnelle aborde un texte par son/ ses thème(s) au niveau macro-sémantique : unités lexicales directement identifiables, et par sa syntaxe (connecteurs argumentatifs articulant les propositions)..
▪L'A.L.S. traite le seul lexique, et à un niveau micro-sémantique, infra-thématique (les "lunettes subjectives" sont indépendantes du monde perçu "objectivement"). Elle décompose les unités lexicales en éléments minimaux, ce qui permet l’analyse d'un contenu subjectif apparemment "irrationnel", mais disposant de sa propre logique..
Cette analyse débouche
▪sur des règles d'annotation en noir et blanc, mais aussi en "fausses couleurs", facilitant la perception du "diagnostic" d'un simple coup d'œil,
▪et sur une objectivation graphique (courbes) de la dimension subjective commune à une famille de textes,
… et d'autre part l'A.L.S. bénéficie, grâce au travail de Vincent Risch, d'une formalisation partielle au moyen des X-logiques (Pierre Siegel).
- Le 03/07/2023 à 14h à St Charles, FRUMAM, 2ème étage
Orateur : Emiliano Lorini (IRIT)
Titre : A Rule-based Modal View of Causal Reasoning
Résumé : We present a novel rule-based semantics for causal reasoning as well as a number of modal languages interpreted over it. They enable us to represent some fundamental concepts in the theory of causality including causal necessity and possibility, interventionist conditionals and Lewisian conditionals. We provide complexity results for the satisfiability checking and model checking problem for these modal languages. Moreover, we study the relationship between our rule-based semantics and the structural equation modeling (SEM) approach to causal reasoning, as well as between our rule-based semantics for causal conditionals and the standard semantics for belief base change.
- Le 05/06/2023 à 10h à St Charles, FRUMAM, 3ème étage
Orateur : Simon Vilmin (LORIA)
Titre : Enumération des modèles maximaux d'une formule de Horn
Résumé : L'énumération des modèles maximaux d'une formule de Horn est un problème difficile qui apparaît implicitement dans plusieurs domaines de l'informatique, allant du data mining à la logique modale. En outre, il généralise la dualisation des fonctions monotones booléennes (ou dualisation des hypergraphes) et se rapporte au problème de dualisation dans les treillis codés par des bases d’implications. Dans cet exposé, je présenterai le problème et quelques résultats connus à son sujet. En général, le problème ne peut être résolu en temps total polynomial à moins que P = NP, ce qui suggère l'étude de cas particuliers. Je m'intéresserai au cas où les clauses négatives de la formule ont taille deux, qui fait écho au codage d'un graphe médian au moyen d'un ordre partiel et d'un graphe de conflit. En tirant parti du lien entre formules de Horn et treillis, j'utiliserai le nombre de Carathéodory (que l'on retrouve notamment en optimisation combinatoire) et la dualisation des hypergraphes pour donner des classes de formules où le problème peut être résolu en temps total (quasi-)polynomial. Je mentionnerai aussi des classes de treillis où le problème est déjà plus dur que la dualisation des hypergraphes, même si les clauses négatives sont de tailles deux.
- Le 15/05/2023 à 14h à St Charles, FRUMAM, 2ème étage
Orateur : Philippe Balbiani (IRIT)
Titre : Unification in non-classical logics
Résumé : Originally motivated by automatic deduction tools, the unification problem in modal logics constitutes today the heart of an active research area in which much remains to be done. Within the context of description logics, the main motivation for investigating the unification problem was to propose new reasoning services in the maintenance of knowledge bases like the elimination of redundancies in the descriptions of concepts. In this talk, we will give a survey of the results on unification in these non-classical logics and we will present some of the open problems whose solution will have a great impact on the future of the area.
- Le 05/05/2023 à 14h30 à St Charles, FRUMAM, 3ème étage
Orateur : Karim Tabia (CRIL)
Titre : Reasoning with Uncertain Inputs in an Imprecise Probabilistic Setting
Résumé : Probability intervals provide an intuitive, powerful and unifying setting for encoding and reasoning with imprecise beliefs. In this presentation, I will talk about some recent work related to reasoning with uncertain information in the framework of probability intervals. Il will first discuss the problem of updating uncertain information specified in the form of probability intervals with new uncertain inputs also expressed as probability intervals. Secondly, I will discuss the inverse problem of probabilistic belief update with uncertain inputs. Finally, I present an application of reasoning with uncertain information to better calibrate the predictive probabilities of machine learning models.
- Le 10/04/2023 à 14h à distance
Orateur : Samuel Pastva (Institute of Science and Technology, Vienna)
Titre : Robust control of Boolean networks in systems biology
Résumé : Boolean networks represent a simple yet scalable model of gene regulation in living organisms. Their ability to mimic biological systems makes them useful when attempting to reprogram the function of biological circuits (e.g. altering cell phenotypes). Depending on the specific biological conditions, this problem comes in many flavours, but is collectively referred to as Boolean network control. Today, even the most detailed Boolean models often contain elements that are either added ad hoc or fine-tuned (arguably over-fitted) to a particular dataset. An alternative is to view these elements as unknown or undetermined, which is more true to the available biological knowledge. However, this introduces uncertainty in the network dynamics which complicates computational analysis and interpretation of results. This talk provides an overview of the various sub-problems of Boolean network control and methods for addressing these problems computationally. It then expands this view to partially specificed Boolean networks and introduces a notion of robust control which allows to safely account for uncertainty in network dynamics. The viability of robust control is explored using a collection of symbolic, BDD-based algorithms.
- Le 03/04/2023 à 14h à distance
Orateur : Maria Boritchev (Orange Innovation, Lannion)
Titre : Logique, compositionnalité, et sémantique pour la modélisation des langues naturelles
Résumé : En linguistique computationnelle, la question de l'accès à la sémantique des textes écrits dans des langues naturelles (dans cet exposé, principalement anglais et un peu français) a été principalement abordée de deux points de vue : un point de vue statistique et un point de vue formel, au sens de la logique mathématique. La première approche tire parti de grandes quantités de données annotées disponibles pour l'anglais et utilise des techniques d'apprentissage automatique pour produire des analyseurs sémantiques qui fonctionnent, en moyenne, assez bien, tout en produisant des résultats parfois absurdes pour des phénomènes linguistiques et logiques tels que la négation et la quantification. La seconde approche est ancrée dans la logique mathématique et dans la compositionnalité, ce qui la rend robuste et bien adaptée à la modélisation des phénomènes linguistiques et logiques dans le langage, mais rend très difficile le passage à l'échelle de cette approche. Plutôt que choisir l'une des deux approches et de s'y tenir, nous aimerions étudier la possibilité de les combiner pour obtenir un analyseur sémantique logiquement robuste et évolutif, aussi adaptable que possible aux spécificités des différentes langues. Le premier pas que nous faisons dans cette direction est l'étude de la capacité de deux architectures de réseaux neuronaux (perceptron multi-couches, MLP, et réseau de neurones récurrent, RNN) à produire des raisonnements compositionnels.
- Le 20/03/2023 à 14h à distance
Orateur : Karim Tabia (Université d'Artois, CRIL)
Titre : Explainable AI: A declarative approach for multi-label classification
Résumé : With the rise of artificial intelligence systems in many areas (autonomous car, smart cities, predictive health, etc.), it becomes essential and critical to understand the decisions and behaviors of an intelligent system. In this presentation, I will first give the big picture of explainable AI and explainability in machine learning. Secondly, the focus will be on declarative approaches that we have proposed to provide both symbolic and numerical explanations for multi-class and multi-label classification problems. The presentation will point out, finally, some problems and challenges facing the AI community in terms of explainable AI.
- Le 06/02/2023 à 13h15 à St Charles, FRUMAM, 3ème étage
Orateur : Marianna Girlando (Université d'Amsterdam)
Titre : Labelled calculi for non-classical logics
Résumé : Labelled sequents enrich the syntax of Gentzen-style calculi by introducing in the language pieces of semantic information, the "labels", which represent elements of possible-world models for the logic.
In this talk we introduce the formalism of labelled sequent calculi, and showcase how modular and analytic labelled systems can be defined for a number of modal and non-classical logics - namely, those whose frame conditions can be expressed by means of geometric axioms, a specific class of first-order formulas. As a case study, we shall present labelled calculi for conditional logics. We will conclude by discussing how labelled calculi could be defined for modal logics with recursive modalities, thus combining the labelled approach with tools from cyclic proof theory.
- Le 05/12/2022 à 14h à Luminy (salle 4.05, Bât. TPR2 au LIS) et en visio
Orateur : Cameron Calk (Post-Doc LIS)
Titre : Coherence via strategic rewriting in higher Kleene algebras
Résumé : Normalisation strategies provide a categorical interpretation of contracting homotopies via confluent and terminating rewriting. In particular, this approach relates standardisation to coherence results in the context of higher dimensional rewriting systems. On the other hand, modal Kleene algebras (MKAs) have provided a description of properties of abstract rewriting systems, and classic (one-dimensional) consistency results have been formalised in this setting.
In this talk, I will recall the notion of higher Kleene algebra, introduced as an extension of MKAs, and which provide a formal setting for reasoning about (higher dimensional) coherence proofs for abstract rewriting systems. In this setting, I will describe how the formalisation of the notion of normalisation strategy leads to a coherence theorem via convergent rewriting. At the end, I will briefly discuss how natural examples of such structures arise from a Jónsson-Tarski duality between n-categories and globular n-quantales.
Lien zoom: https://univ-amu-fr.zoom.us/j/82328232953?pwd=d3d2TkVkY2tkOXdhQzRnSnZaeURGQT09
- Le 17/10/2022 à 14h à St Charles, FRUMAM, 2ème étage
Orateur : Van-Giang Trinh (LIRICA)
Titre : Minimal trap spaces of Boolean models are maximal siphons of their Petri net encoding
Résumé : Boolean modelling of gene regulation but also of other biological systems has had great successes over the last ~20 years. Besides simulation, the analysis of such models is mostly based on attractor computation, since those correspond roughly to observable biological phenotypes. The recent use of trap spaces made a real breakthrough in that field allowing to consider medium-sized models that used to be out of reach. However, with the continuing increase in model-size, the state-of-the-art computation of minimal trap spaces based on prime-implicants shows its limits as there can be a huge number of implicants.
In this research, we present an alternative method to compute minimal trap spaces, and hence complex attractors, of a Boolean model. It replaces the need for prime-implicants with a completely different technique, namely the enumeration of maximal siphons in the Petri net encoding of the original model. After some technical preliminaries, we expose the concrete need for such a method and detail its implementation using Answer Set Programming. We then demonstrate its efficiency and compare it to implicant-based methods on several large Boolean models from the literature.
- Le 03/10/2022 à 14h à St Charles, FRUMAM, 2ème étage
Orateur : Nicola Olivetti (LIRICA)
Titre : A Journey in intuitionistic Modal Logic: normal and non-normal modalities
Résumé : Modal extensions of intuitionistic logic have a long history going back to the work by Fitch
in the 40’ [6]. Two traditions are now consolidated, called respectively Intuitionistic Modal
Logic and Constructive Modal logic. Each of the two has its own motivation and is more
natural than the other from some standpoint. In the former tradition originated by Fischer-
Servi [5] and systematized by Simpson [9], the basic system is IK, whereas in the tradition
of constructive modal logics the two basic systems are Wijesekera’ systems WK [10] and the
system CK by Bellin et als. [1]. Constructive modal logic are non-normal modal logics. In the
classical setting, non-normal modal logics have been studied for a long time for several purposes
(see [2], [8]). The observation that constructive modal logics are non-normal and the interest
in non-normal modalities in itself leads to the question: which are the intuitionistic analog of
classical non-normal modal logic?
It turns out that the framework of intuitionistic non-normal modal logic is richer than the
classical one. In particular different interactions between the two modalities 2 and 3 give
rise to distinct systems; some of them do not have a counterpart in the classical case. The
resulting picture is a lattice of 24 non-normal modal logics with an intuitionistic base each of
them determined by a cut-free sequent calculus.
Similarly to classical non-normal modal logics, all systems of non-normal intuitionistic modal
logic are characterized by a simple neighbourhood semantics. Moreover the neighbourhood
semantics helps to understand also Constructive modal logics CK and WK, as it covers also
these systems.
The interest of the neighbourhood semantics for constructive modal logic can also be justified
from a proof-theoretical perspective, as it witnessed by some recently introduced unprovability
calculi for these logics. In these calculi, each derivation precisely corresponds to one neighbour-
hood countermodel, whereas there is no direct correspondence with relational models. This
fact confirms the usefulness and the naturalness of neighbourhood semantics for analysing in-
tuitionistic modal logics.
[Joint work with Tiziano Dalmonte and Charles Grellois.]
- Le 19/09/2022 à 14h à St Charles, FRUMAM, 2ème étage
Orateur : Giulio Guerrierri (LIRICA)
Titre : Understanding the lambda-calculus via (non-)linearity and rewriting
Résumé : The lambda-calculus is the model of computation underlying functional programming languages and proof assistants. Actually, there are many lambda-calculi, depending on the evaluation mechanism (e.g., call-by-name, call-by-value, call-by-need) and computational features that the calculus aims to model (e.g., plain functional, non-deterministic, probabilistic, infinitary).
The existence of different paradigms is troubling because one apparently needs to study the theory and semantics of each one separately.
We propose a unifying and uniform meta-theory of lambda-calculi, where the study is rooted on a unique core language, the bang calculus, a variant of the lambda-calculus inspired by Girard's linear logic and Levy's Call-by-Push-Value: a bang operator marks which resources can be duplicated or discarded.
The bang calculus subsumes both call-by-name and call-by-value. A property studied in the bang calculus is automatically translated in the corresponding property for the call-by-name lambda-calculus and the call-by-value lambda-calculus, thanks to two different embeddings of the lambda-calculus in the bang calculus. These embeddings are grounded on proof theory, via Curry-Howard: they are an adaptation of Girard's two translations of intuitionistic logic into linear logic.
Advanced computational features are usually obtained by adding new operators to a core language. Instead of studying the operational properties of the extended language from scratch, we propose a simple method to study them modularly: if an operational property holds in the core language and in the new operators separately, then a simple test of good interaction between core language and new operators guarantees that the property lifts to the whole extended language. This approach is first developed in abstract rewriting systems.
- Le 13/07/2022 à 10h à St Charles, FRUMAM, 2ème étage
Orateur : Trinh Van Giang (LIRICA)
Titre : Minimal trap spaces of Boolean models are maximal siphons of their Petri net encoding
Résumé : Boolean modelling of gene regulation but also of other biological systems has had great successes over the last ~20 years. Besides simulation, the analysis of such models is mostly based on attractor computation, since those correspond roughly to observable biological phenotypes. The recent use of trap spaces made a real breakthrough in that field allowing to consider medium-sized models that used to be out of reach. However, with the continuing increase in model-size, the state-of-the-art computation of minimal trap spaces based on prime-implicants shows its limits as there can be a huge number of implicants.
In this research, we present an alternative method to compute minimal trap spaces, and hence complex attractors, of a Boolean model. It replaces the need for prime-implicants with a completely different technique, namely the enumeration of maximal siphons in the Petri net encoding of the original model. After some technical preliminaries, we expose the concrete need for such a method and detail its implementation using Answer Set Programming. We then demonstrate its efficiency and compare it to implicant-based methods on several large Boolean models from the literature.
- Le 13/06/2022 à 14h en visio et à St Charles, FRUMAM, 2ème étage
Orateur : Marianna Girlando (University of Birmingham, UK)
Titre : Cyclic proofs, hypersequents and Transitive Closure Logic
Résumé : Transitive Closure Logic (TCL) enriches the language of first order logic with a recursive operator, expressing the transitive closure of a binary relation. Cyclic sequent calculi proofs for this logic have been introduced in the literature. In this talk, I will present a new cyclic proof system for TCL, which employs hypersequents. The main motivation behind this work is that, differently from sequent TCL proofs, hypersequent proofs for TCL successfully simulate cyclic proofs in the transitive fragment of Propositional Dynamic Logic (PDL+), without using the non-analytic rule of cut.
After introducing the cyclic hypersequent calculus for TCL, I will present two main results. First, I will sketch a proof of soundness for the calculus, which requires a more involved argument than the one used in traditional cyclic proofs. Second, I will show how PDL+ cyclic proofs can be translated into TCL hypersequent proofs. This result allows to lift the completeness result for cyclic PDL+ to the hypersequent calculus.
This talk is based on joint work with Anupam Das.
- Le 30/05/2022 à 14h en visio
Orateur : Raphaëlle Crubillé (LIRICA)
Titre : Sémantique dénotationelle pour les langages de programmation probabilistes
Résumé : La sémantique dénotationelle est un champs de recherche centré sur la construction de modèles mathématiques des langages de programmation, visant à produire des techniques de preuve pour l'équivalence de programmes. Dans les dix dernières années, l'usage des probabilités pour la programmation a été transformé par les développements en programmation statistique et en apprentissage, ce qui a stimulé un intérêt renouvelé pour les méthodes formelles pour des langages probabilistes. Cet exposé présente un bref panorama de la sémantique dénotationelle des langages de programmation probabilistes, en tentant d'expliciter à la fois ses enjeux, ses outils et ses perspectives.
- Le 09/05/2022 à 14h en visio
Orateur : José-Luis Vilchis-Medina (Ecole navale, Brest)
Titre : Autonomous Decision-Making with Incomplete Information and Safety Rules based on Non-Monotonic Reasoning
Résumé : In this seminar, we will present a decision process integrating Non-Monotonic Reasoning (NMR), embedded in a deliberative architecture. The NMR process uses Default Logic to implement goal reasoning, managing partially observable or incomplete information, allowing the design of default behaviours completed by the handling of specific situations, in order to manage the current mission objective as well as safety rules. We illustrate our approach through an application of an underwater robot performing a marine biology mission.
- Le 25/04/2022 à 14h en visio et à St Charles, FRUMAM, 2ème étage
Orateur : Tarek Khaled (LIRICA)
Titre : De la conception d'un système ASP au traitement des données biologiques
Résumé : La programmation par ensemble réponse (Answer Set Programming, ASP) est un paradigme de programmation déclaratif basé sur la sémantique des modèles stables. L'idée derrière ASP est de coder un problème donné comme un ensemble de règles logiques, les modèles stables obtenus à partir du programme logique constituent les solutions du problème initial. La grande expressivité de l'ASP, combinée au nombre croissant d'applications, ont fait de l’ASP un sujet de recherche majeur.
Dans cet exposé, nous présentons la manière dont nous avons exploité une nouvelle sémantique de programme logique pour la conception d'un nouveau système ASP. Cette sémantique nous a donné la possibilité de calculer tous les modèles stables d'un programme logique ainsi que les modèles extra-stables. Ces modèles extra-stables correspondent aux extra-extensions de la nouvelle sémantique qui ne sont pas capturées par la sémantique des modèles stables. La méthode que nous proposons effectue un processus énumératif booléen adapté à l’ ASP et à la nouvelle sémantique. Elle a l’avantage de travailler sur une représentation sous forme de clauses de Horn ayant la même taille que le programme logique d’entrée et travaille à espace constant. En plus, l’énumération est effectuée sur un ensemble restreint de littéraux du programme logique représentant son "strong backdoor".Dans la suite de cet exposé, nous aborderons l'utilisation de programmes logiques et d'ASP pour exprimer et traiter des réseaux de régulation génique considérés comme des réseaux booléens, et nous présenterons des méthodes permettant de détecter tous les attracteurs dans ces réseaux. Les attracteurs sont essentiels dans l'analyse de la dynamique d'un réseau booléen. Ils expliquent pourquoi une cellule particulière peut acquérir des phénotypes spécifiques qui peuvent être transmis sur plusieurs générations. Nous présentons d'abord comment représenter et traiter les réseaux booléens généraux pour les modes de mise à jour synchrone et asynchrone. Ensuite, nous faisons une présentation du cas particulier des réseaux circulaires. Ce dernier cas spécifique joue un rôle essentiel dans les systèmes vivants. Nous avons montré plusieurs propriétés théoriques; en particulier, les attracteurs simples des réseaux de gènes sont représentés par les modèles stables des programmes logiques correspondants et les attracteurs cycliques par leurs modèles extra-stables.
- Le 04/04/2022 à 14h en visio et à St Charles, FRUMAM, 2ème étage
Orateur : Federico Olimpieri (Leeds, UK)
Titre : A categorical approach to resource approximation
Résumé : In recent years, the semantics of linear logic has lead to the development of powerful tools for the analysis of quantitative and qualitative aspects of computation. The notion of resource approximation arose in this context and it consists in the study of the computational behavior of programs via an appropriate kind of approximations. An important example of this is given by linear approximations of lambda-terms. These approximations are resource-sensitive programs, that use exactly once their input during computation. I will present some relevant results in this field from my general perspective rooted in the categorical semantics of programming languages.
Transparents exposé- Le 21/03/2022 à 14h en visio
Orateur : Sophie Paillocher (IBISC, Evry)
Titre : Complétion de modèles partiellement étiquetés pour la logique LTL
Résumé : Un problème important dans le processus de conception et réalisation d’un système informatique
critique est de tester automatiquement si un modèle abstrait donné du système vérifie ou non une
spécification donnée, exprimée dans un langage logique. Ce problème, connu sous le nom de
« Model Checking » est usuellement étudié dans le cadre où le modèle abstrait est entièrement
construit.
Toutefois des scénarios existent où l’information dont l’on dispose à propos du système est
partielle. En effet, le système peut être seulement partiellement construit, ou bien la possibilité
d’observation du système peut être limitée.
Dans ce contexte deux problèmes émergent :
• Peut-on compléter le modèle partiel de façon à obtenir un modèle entièrement construit
respectant une spécification donnée ?
• Peut-on s’assurer que le modèle respectera une spécification donnée, peut importe
l’information manquante?
Ici nous considérerons les modèles comme étant des systèmes de transitions dont les états sont
étiquetés par des informations sur les propositions atomiques. Le cadre logique sera quant-à lui
celui de la logique LTL. De plus nous restreindrons l’information manquante aux seules étiquettes.
La structure du modèle partiel (états et transitions) donné en entrée restera donc inchangée lors de la
complétion.
Nous commencerons par analyser les problèmes de complétion, et en particulier par énoncer des
résultats de complexité. Puis nous présenterons une méthode générale, inspirée de la méthode des
tableaux, afin de les résoudre.
- Le 07/03/2022 à 14h en visio et à St Charles, FRUMAM, 2ème étage
Orateur : Ezgi Iraz SU
Titre : Recent Epistemic Extensions of Equilibrium Logic
Résumé : Answer set programming (ASP) is a well-established paradigm for declarative logic programming and nonmonotonic reasoning. Thanks to its efficient solvers, ASP performs as a successful logic-based problem-solving technique and is especially useful for difficult combinatorial search problems with optimisation.
Here-and-there logic (HT) is a nonclassical (three-valued) monotonic logic which is intermediate between classical logic and intuitionistic logic, and it is closely aligned with ASP:
(i) First, Pearce’s equilibrium logic provides a purely logical framework for ASP. Its semantics, via equilibrium models, is based on a truth-minimisation criterion over HT models. The way of choosing such minimal models among HT models of a formula turns equilibrium logic into a nonmonotonic reasoning formalism.
(ii) Second, HT characterises an important notion of ASP. To spell it out, strong equivalence of ASP programs can be captured by logical equivalence of corresponding HT theories (i.e., sets of HT formulas).
In this talk, we first make a short introduction to the above-mentioned formalisms with a focus on equilibrium logic. Then, we discuss the recent epistemic extensions of equilibrium logic. Among these, the main approach is to integrate (reflexive) autoepistemic logic and equilibrium logic in order to insert the introspective reasoning of the former to the latter. Autoepistemic logic is an important form of nonmonotonic reasoning, allowing a rational agent to reason about her own knowledge. It extends classical propositional logic by an epistemic modal operator. Autoepistemic logic and its reflexive extension are respectively identical to the nonmonotonic variants of the modal logics KD45 and SW5. Their minimal models reflect minimisation of knowledge. Thus, equilibrium models of (reflexive) autoepistemic equilibrium logics respect both knowledge and truth minimality. Today, they are considered the most successful approach among the existing semantics for epistemic ASP; yet the definition of a rigorous and understandable semantics for epistemic operators in the language of ASP is still subject of ongoing research.
- Le 11/02/2022 à 10h en visio (séminaire commun DALGO-LIRICA)
Orateur : Jérémy Ledent
Titre : Simplicial Models for Multi-Agent Epistemic Logic
Résumé : Epistemic Logic is the modal logic of knowledge. It allows to reason about a finite set of agents who can know facts about the world, and about what the other agents know. The traditional Kripke-style semantics for epistemic logic is based on graphs whose vertices represent the possible worlds, and whose edges indicate the agents that cannot distinguish between two worlds. In this talk, I will present an alternative semantics for epistemic logic, based on combinatorial topology. The idea is to replace the Kripke graph by a simplicial complex, allowing for higher-dimensional connectivity between the possible worlds. In fact, every Kripke model can be turned into an equivalent simplicial model, thus uncovering its underlying geometric structure. Our notion of simplicial model is inspired from the "protocol complex" approach to distributed computing. I will show how our framework can be used to analyse distributed computing, where the agents are the processes, and the possible worlds are all the possible executions of the system. In order to prove impossibility results, one must find an epistemic logic formula representing the knowledge that the processes should acquire in order to solve a task; and argue that such knowledge cannot be achieved. This is joint work with Éric Goubault and Sergio Rajsbaum
- Le 10/01/2022 à 14h en visio et à St Charles, FRUMAM, 3ème étage
Orateur : Pierre Clairambault (LIRICA)
Titre : Games with no Winner: an Introduction to Game Semantics
Résumé : How to define rigorously the behaviour of programs under execution?
Program semantics is a necessary first step in order to prove that a program or program transformation is correct, and is of course a preliminary to any theoretical question on programming languages. Game semantics is one of many approaches to program semantics. It represents a program by recording the exchange of observable computational events with its execution environment, phrased as a play in a two-player game that the program plays with its context. In game semantics there is no winner: it is about the journey, not the destination!
Most of my talk will be an introduction to game semantics. I will first focus on some old (but nice) results, namely the so-called « semantic cube » following which game semantics exactly captures the observable behaviour of programs with various computational effects (notably, mutable state and non-local control). I will show how this allows one to prove results about computational effects and their interferences.
Time permitting, I will then talk about some of my own recent work on a game semantics for concurrent programs. The semantics builds on so-called « true concurrency » models such as event structures, and provides a causal description of the behaviour of concurrent programs with rich computational features.
- Le 22/11/2021 à 14h à St Charles, FRUMAM, 3ème étage
Orateur : Erkko Lehtonen (Universidade Nova de Lisboa)
Titre : Clonoids of Boolean functions
Résumé : A clonoid is a class of functions of several arguments that is stable under right and left compositions with fixed clones. In this talk, we discuss first steps towards Post-like complete descriptions of clonoids of Boolean functions. We provide a brief introduction to the topic and present some motivations for studying clonoids.
- Le 28/06/2021 à 11h en visio
Orateur : Van-Giang TRINH (Tokyo, Japan)
Titre : An FVS-based Approach to Attractor Detection in Asynchronous Boolean Networks
Résumé : Boolean Networks (BNs) play a crucial role in modeling and analyzing biological systems. One of the central issues in the analysis of BNs is attractor detection, i.e., identification of all possible attractors. This problem becomes more challenging for large Asynchronous Boolean networks (ABNs) because of the asynchronous and non-deterministic updating scheme. In this research, we formally state and prove several relations between Feedback Vertex Sets (FVSs) and dynamics of BNs. From these relations, we propose an FVS-based method for detecting attractors in ABNs. Our approach relies on the principle of removing arcs in the state transition graph to get a candidate set and the reachability property to filter the candidate set. We formally prove the correctness of the proposed method, and then show its efficiency by conducting experiments on real biological networks and randomly generated N-K networks. The obtained results are very promising, since our method can handle large networks whose sizes are up to 101 without using any network reduction technique. This advanced computation capability can enable biologists to conduct more accurate analysis on large networks, then discover more biological insights.
- Le 31/05/2021 à 14h en visio
Orateur : Ezgi Iraz SU
Titre : Recent Epistemic Extensions of Answer Set Programming
Résumé : Answer set programming (ASP) is a well-established paradigm for declarative programming and nonmonotonic reasoning. Accordingly, it serves as a successful method of problem-solving in knowledge representation and reasoning. However, today it is widely accepted by the logic programming community that ASP requires a more powerful introspective reasoning with the use of epistemic modalities. Although there has been a long-lasting debate among researchers about how to correctly extend ASP with epistemic modal operators, there is still no agreement on a fully satisfactory semantics that is able to offer intuitive results for epistemic logic programs (ELPs). To carry this discussion on a more formal level, Cabalar et al. have recently proposed some candidate structural criteria that successful semantics approaches for ELPs are supposed to satisfy. These are mainly the epistemic extensions of ASP’s well-known principles of constraint monotonicity, splitting, and foundedness.
In this talk, we first give a brief survey of ASP’s recent epistemic extensions, and if time permits, then we also shortly discuss these formal properties.
- Le 26/04/2021 à 14h en visio
Orateur : Wesley Fussner
Titre : Ortholattices résiduels comme nouvelle approche de la logique quantique
Résumé : Cet exposé passe en revue nos récentes tentatives d'étudier le raisonnement quantique à l'aide des outils des logiques sous-structurelles. Nous introduisons les ortholattices résiduels comme des modèles algébriques généralisant les modèles habituels de la logique quantique, et illustrons qu'ils donnent une version `intuitioniste' de la logique quantique dans laquelle la logique quantique classique peut être interprétée par une traduction à double négation.
Séminaires passés (depuis 2016)