Embedded Files
PROGRAM
PROGRAM
Tuesday, 7 November 2023
8.30 - 9.15
Registration
9.15 - 9.30
Opening
9.30 - 10.30
“Fabio Bellifemine” Keynote Speech:
Cristiano CASTELFRANCHI "Eliminativism for AI? No longer ‘cognitive representations’ and ‘minds’ in artificial intelligences?” [Chair Rino Falcone]
10.30 - 11.00
Coffee Break
11.00 - 13.05
Session 1: Learning and Advancements in Multi-Agent Systems [Chair Viviana Mascardi]
11.00 - 11.25 RMLGym: a formal reward machine framework for reinforcement learning. Hisham Unniyankal, Francesco Belardinelli, Angelo Ferrando and Vadim Malvone
11.25 - 11.50 Capturing Recursive Patterns in Neural-Symbolic Reinforcement Learning. Davide Beretta, Stefania Monica and Federico Bergenti
11.50 - 12.15 Curriculum–Based Reinforcement Learning for Pedestrian Simulation: Towards an Explainable Training Process? Giuseppe Vizzari, Daniela Briola and Thomas Cecconello
12.15 - 12.40 Peer-Reviewed Federated Learning. Mattia Passeri, Andrea Agiollo and Andrea Omicini
12.40 - 13.05 Agents in software development architectures. Marco Loaiza, Claudio Savaglio, Niaz Hussain Arijo, Gianluca Aloi, Giancarlo Fortino and Raffaele Gravina
13.05 - 13.35
WOA Steering Committee Meeting
13.35 - 15.00
Lunch Break
15.00 - 16.15
Session 2: Cognition, trust and reputation [Chair Andrea Omicini]
15.00 - 15.25 Agents showing self-disclosure. A preliminary methodological approach. Valeria Seidita, Angelo Maria Pio Sabella and Antonio Chella
15.25 - 15.50 A Strategy to Detect Colluding Groups by Reputation Measures. Attilio Marcianò, Domenico Rosaci and Giuseppe M.L. Sarnè
15.50 - 16.15 Dependence Networks and Trust in Agents Societies: Insights and Practical Implications. Rino Falcone and Alessandro Sapienza
17.45 Visit to a Roman subterranean site... it will be a surprise!
20.00 Social Dinner
Wednesday, 8 November 2023
9.20 - 10.35
Session 3: Theoretical approaches [Chair Alessandro Sapienza]
9.20 - 9.45 n-mates evaluation: a new method to improve the performance of genetic algorithms in heterogeneous multi-agent systems. Paolo Pagliuca and Alessandra Vitanza
9.45 - 10.10 A cognitive approach to model Intelligent Collaboration in Human-Robot Interaction Filippo Cantucci and Rino Falcone
10.10 - 10.35 A timed epistemic logic for formalizing cooperation among groups of agents. Stefania Costantini, Andrea Formisano and Valentina Pitoni
10.35 - 11.00 coffee break
11.00 - 13.05
Session 4: Applicative studies [Chair Filippo Cantucci]
11.00 - 11.25 Agents for Industry 4.0: the Case Study of a Production Cell. Matteo Baldoni, Cristina Baroglio, Valeriano Ditano, Roberto Micalizio and Stefano Tedeschi
11.25 - 11.50 JaKtA: BDI agent-oriented programming in pure Kotlin. Martina Baiardi, Samuele Burattini, Giovanni Ciatto and Danilo Pianini
11.50 - 12.15 A system for tracking patients in the operating room - A pilot study. Mattia Pellegrino, Gianfranco Lombardo, Monica Mordonini, Stefano Cagnoni, Eleonora Bottani, Valentina Bellini, Elena Giovanna Bignami and Agostino Poggi
12.15 - 12.40 Modeling the video game environment: the VideOWL ontology. Simone De Martino, Marianna Nicolosi-Asmundo, Stefano Angelo Rizzo and Daniele Francesco Santamaria
12.40 - 13.05 A Declarative C++ Agent Platform for Agent-based Edge Computing. Fabrizio Messina, Corrado Santoro and Federico Fausto Santoro
13.05 - 14.40
Lunch Break
14.40 - 15.30
Session 5: The role of Explainability [Chair Giuseppe Vizzari]
14.40 - 15.05 Towards a Multi-Level Explainability Framework for Engineering and Understanding BDI Agent Systems. Elena Yan, Samuele Burattini, Jomi Fred Hübner and Alessandro Ricci
15.05 - 15.30 Unlocking Insights and Trust: The Value of Explainable Clustering Algorithms for Cognitive Agents Federico Sabbatini and Roberta Calegari
15.30 - 16. 00
Coffee break
16.00 - 18.00
Roundtable: Progress and Complexity of AI: A Cross-Sector Exploration
Participants: Mario DE CARO (Università Roma3, Dipartimento di Filosofia), Chiara GALLESE (Università di Torino, Dipartimento di Giurisprudenza) , Dario GUARASCIO (Università Sapienza, Dipartimento di Diritto ed Economia delle attività produttive) [Discussant Rino Falcone]
18.00 - 18.15
Closing
The social dinner
The social dinner
The social dinner will be held on the 7th at 8.30 pm at the Hostaria da Bruno restaurant, Via Varese, 29.
The guided tour
The guided tour
During the evening of Tuesday 7, the organizing committee has the pleasure of offering a guided tour in a subterranean site of the center. Further details will be given on site.
INVITED SPEAKER
INVITED SPEAKER
CRISTIANO CASTELFRANCHI
TITLE: Eliminativism for AI? No longer ‘cognitive representations’ and ‘minds’ in artificial intelligences?
ABSTRACT: We will reconstruct the “eliminativistic” attack to “Cognitive Science” aimed at eliminating notions and models of “mental representations” (“beliefs”, “intentions”, reasoning, ..) and then “mind”. Is now this attack not just in psychology and CogSc but will it propagate - thanks to the “Generative AI” approach - in AI, against the AI models of “intelligence”, of autonomous “Agents”, of their reciprocal “influencing” based on changing the other’s beliefs and goals; of their cooperation based on “mind reading” and goal-adoption. “Cognitive modeling” is on the one side a crucial “scientific” (not just technical) contribution of AI to CogSc, on the other side it’s a fundamental approach/instrument for building “artificial” intelligences, Agents, and hybrid societies. Moreover “transparency” - a crucial condition for human trust in Ags and robots - is not transparency of the underlying “algorithms”; it’s transparency (i) of the hidden “values” and interests; and (ii) of the “minds” of the agent (explainability): the “reasons” of its choice and behavior; its assumptions/beliefs and its goals. That is, “mind reading”. We need to build real minds/intelligences not simply to fake them; “anthropomorphized” machines.
SHORT BIO: Cristiano Castelfranchi is research associate at the Institute of Cognitive Science and Technology, being Director of the Institute in the years 2002-2011. His main research interests concern the study of social norms, trust, and emotions, which resulted in breakthroughs in each of these areas. His emphasis on operational, theory-driven conceptual notions led to collaborations with computer scientists, including early figures in artificial intelligence (AI). In the 1980s, he worked extensively on computational linguistics and later became a key figure in the development of the multi-agent systems approach to Distributed Artificial Intelligence. Recognizing both the strengths and weaknesses of agent-based simulation in analyzing social phenomena, Castelfranchi advocated the need for a robust theoretical foundation for key notions in multi-agent systems (powers, dependence network, autonomy, cooperation, influence, commitment, norms, …) . This approach significantly influenced the development of the field. Castelfranchi also recognized the potential relevance of this methodology for understanding human cognition and society. Castelfranchi's theoretical approach aimed to establish an operational notion of goals, distinct from the vaguer concept of "motivation" prevalent in cognitive and social psychology at the time. He defined goals as anticipatory representations of world-states capable of guiding an agent's behavior, which proved to be invaluable in understanding language and social phenomena, including complex emotions. He systematically investigated the cognitive mediators of social phenomena and framed society as emerging from the interaction of cognitive agents and cognition as being shaped by social interaction. This framework gained significance in computer science, social psychology, economics, and philosophy. In his last decade of scientific activity, Castelfranchi's research focused on the role of society in shaping cognitive processes, emphasizing the centrality of action-control in cognition. His contributions advanced the understanding of cognitive systems, particularly in the domains of anticipatory mechanisms, goal-oriented action-control, and autonomous behavior.
Page updated
Google Sites
Report abuse