Wednesdays@DEI: Talks, 17-05-2023

Autor e vínculos: André Martins, Professor DEI/DEEC

Título: I apologize for any confusion, I made a mistake in my previous response

Abstract: In this talk, I will discuss some of the open problems in artificial intelligence (AI) and natural language processing (NLP) and the planned research underlying my forthcoming ERC project DECOLLAGE (Deep Cognition Learning for Language Generation). Large-scale language models, such as the ones from the GPT family, have led to impressive results in many NLP tasks, exhibiting transfer and few-shot learning capabilities. When interacting with such systems, users commonly find them capable of reasoning, planning, and explaining their decisions, often in convincing ways. However, despite the enormous advances in the last years, I will argue that current deep learning models for NLP are still very limited in fundamental ways and many important ingredients are still missing to achieve a satisfactory level of "intelligence". I will discuss some of these limitations and I will argue that they partly stem from their monolithic architectures, which are good for perception, but unsuitable for tasks requiring higher-level cognition. I will then describe briefly how the DECOLLAGE project will attack these fundamental problems by bringing together tools and ideas from machine learning, sparse modeling, information theory, and cognitive science, in an interdisciplinary approach.