21–22 Nov 2024
Mercure Budapest Castle Hill Hotel
Europe/Budapest timezone

Session

Artificial Intelligence

21 Nov 2024, 09:30
Mercure Budapest Castle Hill Hotel

Mercure Budapest Castle Hill Hotel

H-1013 Budapest, Krisztina Körút 41-43 Tel.: +36 1 488-810

Conveners

Artificial Intelligence

  • Peter Levai (WIGNER RCP)

Artificial Intelligence

  • Gergely Barnafoldi (Wigner RCP RMI of the Hungarian Academy of Sciences)

Presentation materials

There are no materials yet.

  1. Dr Eszter Udvary (Budapest University of Technology and Economics)
    21/11/2024, 09:30
  2. Gábor Érdi-Krausz (HUN-REN SZTAKI)
    21/11/2024, 09:50
  3. Antal Jakovác (WIGNER RCP)
    21/11/2024, 10:25

    Mainstream artificial intelligence (AI) solutions, while achieving considerable success in areas such as classification, generation, and natural language understanding, still face several notorious, long-standing challenges. These include unexpected classification errors, hallucinations in generative models, and catastrophic forgetting, among others.

    Addressing these issues requires...

    Go to contribution page
  4. Péter Antal (Budapest University of Technology and Economics)
    21/11/2024, 11:30

    Large language models (LLMs) face challenges with complex
    probabilistic, causal, and counterfactual reasoning, yet they
    demonstrate promising capabilities in the automated construction of both
    qualitative and quantitative probabilistic and causal models. This talk
    provides an overview of current benchmarks for causal reasoning in LLMs,
    methods for extracting knowledge from LLMs,...

    Go to contribution page
  5. Hunor István Lukács (Eötvös University)
    21/11/2024, 12:00
  6. Andras Telcs
    21/11/2024, 12:30

    Temporally evolving systems are typically modeled by dynamic equations. A
    key challenge in accurate modeling is understanding the causal relationships between subsystems, as well as identifying the presence and influence of unobserved hidden drivers on the observed dynamics. This paper presents a unified method capable of identifying fundamental causal relationships between pairs of systems,...

    Go to contribution page
Building timetable...