Artificial Intelligence Trento Journal Club

Open discussions and lightning talks on artificial intelligence topics

The AIT Journal Club is an open-science initiative aiming to create a friendly environment to discuss and speculate about the recent research trends in Artificial Intelligence, Machine Learning and Computer Science. It is not a formal occasion either! We will have a lightning talk at each meeting to introduce the topic and get the conversation going, leading to having a drink together after the event! It is open to all interested PhDs, students, researchers, (and professors!) who don't mind spending an evening talking about science and making new friends.

August 4th, 2023 - 17:00

FabLab UniTrento, DISI Department, Via Sommarive 9, Povo, Trento

Learning Identity Effects with Graph Neural Networks

Giuseppe Alessio D'inverno (University of Siena)

Graph Neural Networks (GNNs) have emerged in the past years as a powerful tool to learn tasks on a wide range of graph domains in a data-driven fashion; among all the proposed models, the so-called Message-Passing GNNs (MP-GNNs) have gained more and more popularity for their intuitive formulation, strictly linked with the Weisfeiler-Lehman (WL) test for graph isomorphism, which they have been proven equivalent to. From a theoretical point of view, MP-GNNs have been shown to be universal approximators and their generalization capabilities have been recently investigated for MP-GNNs, under a variety of generalization measures (VC-dimension, Rademacher complexity, etc).The aim of our work is to show the potential and practical limitations on the generalization capabilities of MP-GNNs in their (in)ability to learn the so-called identity effects, i.e. the task of determining if an object is composed of two identical patterns or not. We analyze two case studies: two-letters words, extending the existing results for MLPs and RNNs presented in [1], for which it is shown that, using SGD training, the neural network is not able to generalize to unseen patterns, since the training set is invariant to orthogonal transformations; dicyclic graphs, i.e. graphs composed by two cycles of any length, for which we present positive theoretical results, partially supported by numerical results. The theoretical analysis is supported by an extensive experimental study. [1] Brugiapaglia, Simone, M. Liu, and Paul Tupper. Invariance, encodings, and generalization: learning identity effects with neural networks. Neural Computation 34.8 (2022): 1756-1789.

Are you curious to know more?
Join our Telegram channel for updates on the following events!

👉 Join now! 👈

Past events

March 31st, 2023

Communicating between latent spaces with limited semantic correspondence

Luca Moschella - Website (Sapienza University of Rome)

Irene Cannistraci - Website (Sapienza University of Rome)

[] [Recording] [Slides]

Neural architectures can leverage relative representations to guarantee, in practice, invariance to latent isometries and rescalings, effectively enabling latent space communication: from zero-shot model stitching to latent space comparison between diverse settings, architectures, and data modalities. Moreover, an optimization-based method is introduced to discover new parallel anchors from a limited number of seeds, allowing for the discovery of unknown semantic correspondences between different domains, alignment of their relative spaces, and achieving competitive results in several tasks.

March 3rd, 2023

Modern Probabilistic Logic Programming

Pedro Zuidberg Dos Martires - Website (University of Örebro)

[] [Recording] [Slides]

Probabilistic logic programming is a powerful tool that combines three key elements of modern artificial intelligence: logical reasoning, probability theory, and programming languages. In this presentation, I will provide an introduction to modern probabilistic logic programming, starting from the basics. We will begin by discussing vanilla logic reasoning and then progress to probabilistic logic reasoning. We will then move on to the generalization of probabilistic logics to probabilistic logic programming and examine modern extensions. Finally, we will see how such a language can be applied in the domain of cognitive robotics.

February 17th, 2023

Current state of self-driving cars and new approaches coming from cognitive science

Alice Plebe - Website (University of Trento)

[] [Recording] [Slides]

Autonomous vehicles promise to revolutionize society and improve the daily life of many. After decades of slow-paced progress, recent advances in AI are bringing this promise close to reality more than ever. However, the reality we are witnessing is not as bright as expected. It is becoming evident that autonomous vehicles are incapable of the complex reasoning necessary for driving, in situations where humans are still far more performing. That is why we are exploring new methods beyond the classic engineering approach, by looking for inspiration from the only existing intelligent being able to drive: the human. Cognitive mechanisms and brain components involved in driving can inspire new approaches with the potential to finally shorten the gap between artificial drivers and human drivers.

February 3rd, 2023

Global Explainability of GNNs via Logic Combination of Learned Concepts

Steve Azzolin (University of Trento)

[] [Recording] [Slides]

While instance-level explanation of Graph Neural Networks (GNNs) is a well-studied problem with plenty of approaches being developed, providing a global explanation for the behaviour of a GNN is much less explored, despite its potential in interpretability and debugging. Existing solutions either simply list local explanations for a given class, or generate a synthetic prototypical graph with maximal score for a given class, completely missing any combinatorial aspect that the GNN could have learned. In the work that will be presented, we propose GLGExplainer (Global Logic-based GNN Explainer), the first Global Explainer capable of generating explanations as arbitrary Boolean combinations of learned graphical concepts.

January 13th, 2023

Impact of Domain Changes on Task-oriented Dialogue Systems

Tiziano Labruna (Fondazione Bruno Kessler and Free University of Bozen)

[] [Recording] [Slides]

Nowadays conversational systems work quite well when it comes to solve specific tasks, such as restaurant bookings or train ticket purchasing. A problem that every such system has to face, however, is what happens when the world changes, so that the task-specific domain is modified: a restaurant adds a new type of food, a new train station is added, etc. This work demonstrates how the models of a dialogue system degrade their performances in a situation of changes, and proposes some solutions for overcoming this problem.

November 18th, 2022

Some ideas on collaborative learning algorithms

Gabriele Santin - Website (Fondazione Bruno Kessler (FBK))

[] [Recording] [Slides]

Collaborative and Federated Leaning are emerging approaches to manage cooperation between a group of agents for the solution of Machine Learning tasks, with the goal of improving each agent’s performance without disclosing any data. In this talk we present a recently introduced algorithmic approach and some preliminary attempts to generalize the method and understand the theoretical mechanism behind its functioning. In particular, in the first part we will review and discuss a novel algorithmic architecture based on simple Random Forests, designed to the particular case of Anomaly Detection (or classification of rare events), a setting where typical applications often comprise data with sensible information, but where the scarcity of anomalous examples encourages collaboration. In the second part of the presentation we will review some initial ideas to extend this approach to kernel methods, and discuss some interesting phenomena that emerge in this collaborative setting.

June 24, 2022

Graph-based Generative Face Anonymization with Pose Preservation

Nicola Dall’Asen - Website (University of Pisa and University of Trento)

[] [Recording] [Slides]

In the era of deep learning, the availability of large-scale data has undoubtedly brought technological advances. However, the same fact has also fostered the growing concern regarding privacy issues. Visual privacy preservation is mainly achieved via video redaction methods by obfuscating the personally identifiable information (PII) of a data subject, whose face is often the most identity-informative part. We propose AnonyGAN, a GAN-based solution for face anonymization which replaces the visual information corresponding to a source identity with a condition identity provided as any single image. To maintain the geometric attributes of the source face, i.e., the facial pose and expression, and to promote a more natural face generation, we propose to exploit a Bipartite Graph to explicitly model the relations between the facial landmarks of the source identity and the ones of the condition identity through a deep model.

June 10, 2022

Interactive Label Cleaning with Example-based Explanations

Andrea Bontempelli (University of Trento)

[] [Recording] [Slides]

We tackle sequential learning under label noise in applications where a human supervisor can be queried to relabel suspicious examples. Existing approaches only relabel incoming examples that look suspicious to the model. As a consequence, those mislabeled examples that elude (or don't undergo) this cleaning step end up tainting the training data and the model with no further chance of being cleaned. We propose CINCER, a novel approach that cleans both new and past data by identifying pairs of mutually incompatible examples. Whenever it detects a suspicious example, CINCER identifies a counter-example in the training set that -according to the model- is maximally incompatible with the suspicious example, and asks the annotator to relabel either or both examples, resolving this possible inconsistency. The counter-examples are chosen to be maximally incompatible, so to serve as explanations of the model's suspicion, and highly influential, so to convey as much information as possible if relabeled. CINCER achieves this by leveraging an efficient and robust approximation of influence functions based on the Fisher information matrix (FIM). Our extensive empirical evaluation shows that clarifying the reasons behind the model's suspicions by cleaning the counter-examples helps in acquiring substantially better data and models, especially when paired with our FIM approximation.

May 27, 2022

Can We Use Pretrained Language Models for Personal Response Generation?

Seyed Mahed Mousavi - Website (University of Trento)

[] [Recording] [Slides]

There is a growing research interest in developing personalized conversational artificial intelligence (ConvAI). Such systems should understand the user inputs which encompass personal life-events, emotions, and thoughts; and generate an appropriate response in a coherent manner from the opening of the conversation through successive turns until its closure. While deep end-to-end models have been deployed for response generation in task-based and chit-chat settings, such models are known to suffer from inappropriate and generic responses. In this talk, we will discuss the concepts and approaches to develop dialogue systems, the challenges to personalize such systems, and have an insight into an on-going research on grounded Personal Dialogue Agents.

May 13, 2022

Towards Reliable Hybrid Human-Machine Classification Services

Burcu Sayin Günel (University of Trento)

[] [Recording] [Slides]

Abstract: Hybrid classification services are online services that combine machine learning (ML) and humans - either crowd workers or experts - to achieve a classification objective, from relatively simple ones such as deriving the sentiment of a text to more complex ones such as medical diagnoses. In this talk, I will present our approach toward a science for hybrid classification services; discussing key concepts, challenges, and architectures. I will then focus on a central aspect, that of ML calibration, and how it can be achieved with crowdsourced labels.

April 29, 2022

Learning and reasoning with probabilistic satisfiability modulo theories

Paolo Morettin - Website (KU Leuven)

[] [Recording] [Slides]

Recently, probabilistic Satisfiability Modulo Theories has emerged as a very expressive formalism for modelling complex distributions over continuous and discrete variables. By encoding the probability density function over an SMT formula with piecewise polynomials, any joint probability distribution can be represented with arbitrary precision, granting unprecedented flexibility and enabling probabilistic reasoning with both algebraic and logical constraints. In this talk, I will introduce the core concepts, challenges and applications of probabilistic SMT in the fields of machine learning and quantitative formal verification.

April 8, 2022

A Neuro-Symbolic Approach to Structured Event Recognition

Gianluca Apriceno (Fondazione Bruno Kessler and University of Trento)

[] [Recording] [Slides]

Events are structured entities with multiple components: the event type, the participants with their roles, the outcome, the sub-events etc. A fully end-to-end approach for event recognition from raw data sequence, therefore, should also solve a number of simpler tasks like recognizing the objects involved in the events and their roles, the outcome of the events as well as the sub-events. Ontological knowledge about event structure, specified in logic languages, could be very useful to solve the aforementioned challenges. However, the majority of successful approaches in event recognition from raw data are based on purely neural approaches (mainly recurrent neural networks), with limited, if any, support for background knowledge. These approaches typically require large training sets with detailed annotations at the different levels in which recognition can be decomposed (e.g., video annotated with object bounding boxes, object roles, events and sub-events). In this paper, we propose a neuro-symbolic approach for structured event recognition from raw data that uses “shallow” annotation on the high-level events and exploits background knowledge to propagate this supervision to simpler tasks such as object classification. We develop a prototype of the approach and compare it with a purely neural solution based on recurrent neural networks, showing the higher capability of solving both the event recognition task and the simpler task of object classification, as well as the ability to generalize to events with unseen outcomes.

March 25, 2022

Higher-Order Networks and Motif Analysis in Hypergraphs

Quintino Francesco Lotito - Website (University of Trento)

[] [Recording] [Slides]

Over the last two decades, networks have emerged as a powerful tool to analyze the complex topology of interacting systems. From social networks to the brain, several systems have been represented as a collection of nodes and links, encoding dyadic interactions among pairs of units. Yet, growing empirical evidence is now suggesting that a large number of such interactions are not limited to pairs, but rather occur in larger groups. In this seminar, we will discuss how more sophisticated mathematical frameworks such as the hypergraphs can enhance our modeling capabilities for systems involving higher-order interactions. We will see that dealing with such complex structures requires new algorithms to cope with more computationally difficult problems, and new tools and generalizations of classic network ideas to fully exploit the improvements in the expressive power. In the last part of the talk, we will focus on the specific problem of higher-order motif analysis. Higher-order network motifs are defined as statistically over-expressed connected subgraphs of a given number of nodes, which can be connected by higher-order interactions of arbitrary order. We will show how they are able to characterize the local structure of hypergraphs and extract fingerprints at the network microscale of higher-order real-world systems. Moreover, we will discuss the problem from an algorithmic perspective, investigating also some real-world applications. Finally, we will talk about open challenges and possible future directions.

March 11, 2022

Egocentric Temporal Motifs networks

Antonio Longa - Website (Fondazione Bruno Kessler and University of Trento)

[] [Recording] [Slides]

Temporal graphs are structures which model relational data between entities that change over time. Due to the complex structure of data, mining statistically significant temporal subgraphs, also known as temporal motifs, is a challenging task. In this work, we present an efficient technique for extracting temporal motifs in temporal networks. Our method is based on the novel notion of egocentric temporal neighborhoods, namely multi-layer structures centered on an ego node. Each temporal layer of the structure consists of the first-order neighborhood of the ego node, and corresponding nodes in sequential layers are connected by an edge. The strength of this approach lies in the possibility of encoding these structures into a unique bit vector, thus bypassing the problem of graph isomorphism in searching for temporal motifs. This allows our algorithm to mine substantially larger motifs with respect to alternative approaches. Furthermore, by bringing the focus on the temporal dynamics of the interactions of a specific node, our model allows to mine temporal motifs which are visibly interpretable. Experiments on a number of complex networks of social interactions confirm the advantage of the proposed approach over alternative non-egocentric solutions. The egocentric procedure is indeed more efficient in revealing similarities and discrepancies among different social environments, independently of the different technologies used to collect data, which instead affect standard non-egocentric measures.

February 25, 2022

Algorithmic Recourse and Explainable Counterfactual Interventions

Giovanni De Toni - Website (European Laboratory for Learning and Intelligent Systems (ELLIS))

[] [Recording] [Slides]

'Algorithmic recourse' is defined as the ability to provide actionable feedback to unfairly treated users to overturn the decision made by automated decision systems. This feature has become critical for modern machine learning systems that we use to make decisions in several areas of our lives. Counterfactual interventions are a powerful tool to suggest to affected users which actions they have to perform to change the outcome of a black-box model. They give us the intuition of what the state of the world would have been if we had behaved differently. In this talk, we will provide an overview of modern techniques to ensure algorithmic recourse through the generation of explainable counterfactual interventions.

February 11, 2022

Causal Representation Learning

Emanuele Marconato (University of Pisa and University of Trento)

[] [Recording] [Slides]

The two fields of Causality and Machine Learning have a long history and until now they developed separately. In this talk, I will discuss past and recent progress in both fields and envisage a possible connection, precisely the new research direction of Causal Representation Learning. I will focus on the conceptual development of Representation Learning, i.e. learning high-level variables from low-level observations, and motivate that future AI challenges can be addressed leveraging Causality, in particular for transfer learning and for generalizing out of distribution.