Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
La société comme savoir pragmatique des acteurs
No full textInternational audienceEntretien avec Claire Judde de Larivièr
COUPURES D'UN GROUPE ABELIEN TOTALEMENT ORDONNE
No full textIn this article, we study the cuts of a totally ordered abelian group Γ. We begin by recalling some results on ordered sets I and on the associated sets IS(I) and FS(I) of initial and final segments of I.For a totally ordered set I we review the notion of an I-structure defined on a module over a ring R, and the definition of the Hahn product of a family of R-modules indexed by I.The set Cv(Γ)of convex subgroups of a totally ordered group Γ is also a totally ordered set, canonically isomorphic to the set of cuts of the subset Pr(Γ)of principal convex subgroups. One of the first results is then to equip the group Γ with an I-structure where I is the set Pr(Γ) endowed with the opposite order.We associate a convex subgroup with every cut of the group Γ, and conversely, we can associate a family of cuts with every convex subgroup of Γ. It is by looking at these subgroups, and the I-structure of Γ that we can obtain a classification of the different types of cuts.</div
Horn maps of holomorphic functions locally pseudo-conjugate on their parabolic basins
No full text70 pages, 11 figure
Hierarchical gating of aggression by co-transmitting VPM4 neurons in Drosophila melanogaster
No full textInternational audienc
Comparative assessment of hemolymph composition and bioactivity in different bee species
No full textInternational audienceHemolymph, as mediator of immune responses and nutrient circulation, can be used as physiological marker of an insect’s health, environmental quality or ecological adaptations. Recent studies reported intraspecific variation in protein contents and biological activities of the hemolymph of honey bees related to their diet. Here we measured interspecific variation in three common bee species in the Mediterranean Basin with contrasting ecologies: Apis mellifera, Megachile sicula, and Xylocopa pubescens. Despite all the bees were collected in the same area, we found important quantitative and qualitative variations of hemolymph extracts across species. Samples of A. mellifera and M. sicula had much higher protein concentration, anticancer, antimicrobial and antioxidant activities than samples of X. pubescens. This descriptive study suggests life history traits of bee species have strong influences on their hemolymph properties and call for future large scale comparative analyses across more species and geographical areas
HyPlan: Hybrid Learning-Assisted Planning Under Uncertainty for Safe Autonomous Driving
No full textWe present a novel hybrid learning-assisted planning method, named HyPlan, for solving the collision-free navigation problem for self-driving cars in partially observable traffic environments. HyPlan combines methods for multi-agent behavior prediction, deep reinforcement learning with proximal policy optimization and approximated online POMDP planning with heuristic confidence-based vertical pruning to reduce its execution time without compromising safety of driving. Our experimental performance analysis on the CARLA-CTS2 benchmark of critical traffic scenarios with pedestrians revealed that HyPlan may navigate safer than selected relevant baselines and perform significantly faster than considered alternative online POMDP planners
Policy Safety Testing in Non-Deterministic Planning: Fuzzing, Test Oracles, Fault Analysis
No full textInternational audienceRecent work has introduced methodology for testing learned action policies in AI Planning, aiming to effectively identify bug states where policy behavior is sub-optimal. While this work focused on cost-optimality in classical planning, here we apply the core ideas to safety testing in planning with initial-state and action-outcome non-determinism. We cover the entire testing pipeline, introducing fuzzing algorithms to find unsafe policy runs, as well as test oracles to identify bugs where such unsafe behavior could be avoided. Going beyond the previous framework, we introduce a final step to the pipeline, identifying faults which we define to be specific policy decisions – state/action pairs – transitioning from a safe state (where a safe policy exists) to an unsafe state (where no such policy exists). We adapt a range of known algorithms for these purposes, including also approximate ones bounding the number of times we are allowed to diverge from the learned policy. We run comprehensive experiments evaluating each part of our pipeline. Key takeaways are that safety testing can be quite cheap, in contrast to cost-optimality testing; and that variants of Tarjan’s algorithm tend to be highly effective for this purpose
Agnelage d’automne dans le Sud de la France: implications pour l’acquisation et la gestion des ressources au Néolithique
No full textInternational audienceL’organisation des tâches agropastorales et la disponibilité saisonnière des productions animales sont étroitement liées au calendrier de reproduction des troupeaux domestiques. Au Néolithique, les sociétés agropastorales ont dû composer saisonnièrement avec les contraintes environnementales et la physiologie animale. Sous les latitudes tempérées, les brebis alternent annuellement entre une période de fertilité et un repos sexuel, un cycle principalement calé sur la photopériode. Dans le bassin méditerranéen, le cadre climatique, ainsi que l’existence de races ovines méditerranéennes ayant une forte capacité au désaisonnement, ont permis l’émergence de l’agnelage d’automne, contrastant avec le modèle dominant des mises bas printanières en Europe tempérée. Pour reconstuire la saison de naissance des moutons au Néolithique, nous nous appuyons sur l’analyse séquentielle des isotopes stable de l’oxygène (δ18O) dans l’émail dentaire. Des agnelages d’automne ont déjà été observés dès le 6e millénaire avant notre ère dans le sud de la France et en Espagne. Dans le cadre du programme AUTUMN LAMBS, nous élargissons les recherches dans le Sud de la France, tant sur le plan géographique que chronologique afin de mieux comprendre les conditions socio-économiques et paléoenvironnementales ayant conduit à l’émergence et la perduration de ces systèmes d’élevage. Nous présenterons les nouveaux résultats de l’analyse de huit contextes du Sud de la France datés du 6e au 4e millénaires avant notre ère (Montélimar-Daurelle, Grotte Lombard, Les Bréguières, Montou, Mourre de la Barque, Pendimoun, Pertus II, et le Taï), ainsi que de trois contextes d’Europe continentale du 4e millénaires (Arbon Bleiche 3, Chalain 19, Clairvaux XIV). Nos résultats confirment la présence des naissances automnales en Méditerranée dans le Cardial et leur persistance sur la diachronie dans des complexes chronoculturels du VBQ, Chasséen et Ferrières. Ces agnelages d’automne confirment la capacité physiologique des moutons néolithiques au désaisonnement, ainsi que la maîtrise des interactions socio-sexuelles au sein du troupeau par les éleveurs. De plus, ces résultats reflètent un environnement propice, offrant des ressources fourragères suffisantes pour soutenir la lactation en automne, typique du climat méditerranéen. La combinaison d’un agnelage automnal et printanier permettait aux éleveurs de disposer de produits animaux, notamment du lait, sur un temps plus long de l’année, leur offrant ainsi un avantage significatif dans l’acquisition des ressources
A theory of injection-based vulnerabilities in formal grammars
No full textInternational audienceMany systems are controlled via commands built upon user inputs. For systems that deal with structured commands, such as SQL queries, XML documents, or network messages, such commands are generally constructed in a "fill-in-the-blank" fashion: the user input is concatenated with a fixed part written by the developer (the template). However, the user input can be crafted to modify the command's semantics intended by the developer and lead to the system's malicious usages. Such an attack, called an injection-based attack, is considered one of the most severe threat to web applications. Solutions to prevent such vulnerabilities exist but are generally ad hoc and rely on the developer's expertise and diligence. Our approach addresses these vulnerabilities from the formal language theory's point of view. We formally define two new security properties. The first one, "intent-equivalence", guarantees that a developer's template cannot lead to malicious injections. The second one, "intent-security", guarantees that every possible template is intent-equivalent, and therefore that the programming language itself is secure. We thoroughly analyze the decidability of these properties for the most common grammar classes. We conclude by highlighting the technical implications of these results for various settings and tools
