Search results

Abstract Diagrammatic Monte Carlo (DiagMC) is a versatile numerical technique capable of solving strongly correlated fermion systems in a controlled and accurate way. The core idea behind the technique is to sum all connected Feynman diagrams in a …

Abstract The real-time dynamics of interacting quantum systems remains a major challenge in computational quantum physics. Surprisingly, high-order perturbative expansions have recently emerged as a promising approach to address this question, even in …

Abstract In this seminar, I will introduce the diagrammatic Monte Carlo method and discuss its application to the two-dimensional Hubbard model at finite temperature. The results obtained through this approach are controlled and, importantly, address the …

Two letters from and to Albrecht Goetze and an image of a bronze standard from Alaca Höyük found during the 1930s Silvia Alaura has been invited by the assembly of the Collège de France on the proposal of Prof. Dominique Charpin, Mesopotamian Civilization …

Abstract AI-driven decision-making is often evaluated in terms of individual human judgment - are algorithms faster, fairer or less biased than humans? In this talk, I challenge this framework, arguing that the most appropriate class of comparison for …

Abstract The aim will be to reflect on the opacity of AI models and how so-called "explainable" approaches (XAI), often post-hoc, could be improved by shifting the focus from causal explanation to explanation by …

Abstract The explicability of artificial intelligence (AI) is often presented as essential for the adoption of these systems by physicians. However, its usual approach has two limitations: on the one hand, the lack of anchorage in concrete professional …

Naama Friedmann has been invited by the assembly of the Collège de France, at the suggestion of Professors Stanislas Dehaene, Experimental Cognitive Psychology Chair , and Luigi Rizzi, General Linguistics Chair . She will give a series of four lectures in …

Abstract Recent applications of music-generating AI can now generate pieces of different musical genres with unexpected levels of expressiveness. An old and common criticism of algorithmically generated music stems from its supposed inability to express …

Abstract Generative AIs today offer textual, musical or pictorial productions of "artistic" content, which are used by Internet users and professional artists alike. But beyond the new tool on offer here, with its advantages, biases and risks, is there …

Abstract What does AI mean for aesthetics? This paper will look at the consequences of AI-enhanced artistic practices on our relationship to art and its philosophy, placing them in the history of modern art and its debates. It will also look at how art …

Abstract The classic epistemological questions revolving around the notions of trust, dependence and expertise in the acquisition of knowledge are today taking on a new twist with the emergence of what are sometimes called "epistemic technologies" …

Abstract After presenting what I call the " limits of reason " framework and locating AI beyond the upper limit of human reason, I will examine whether the verbal output of AI systems can be considered a form of testimony and whether this forces a …

Abstract AIs are increasingly involved in the information processing processes from which we elaborate our beliefs. Under these conditions, it is legitimate to ask under what circumstances it is legitimate to consider that such processes are reliable and …

Abstract Starting from the idea that an epistemological shift from the identification of causes to spaces of massive data saturated with strong correlations characterizes both most of the algorithms encircling our lives - such as recommendation algorithms …

Abstract The logic of higher categories (or ∞-categories) is a variation of type theory, which is homotopic in nature and in which the notion of category is the primitive concept - the one that is never defined! An adequate generalization of Voevodsky's …

Abstract The linguistic and other productions of certain artificial intelligences are becoming increasingly similar to human productions. In this context, it often seems appropriate to attribute intentional states to these programs. Nevertheless, such an …

Abstract In this talk, we will discuss the current state of the formalization of modern number theory in mathlib, the mathematical library of Lean. We will highlight recent advancements, key challenges that have been addressed, and the broader …

Abstract The latest advances in generative AI are based on a technique called "Chain of Thought" (CoT), which involves making the machine "think" before responding. This was originally a strategy used in queries addressed to classic Large Language Models …

Abstract Large Language Models (LLMs) seem capable of sustaining discussion on all kinds of topics. There are three possible reactions to this observation. We can attribute to LLMs a form of intelligence, which includes a certain understanding of the …