LASIGE Talks are fortnightly/monthly events to publicize recently distinguished publications or ongoing cutting-edge work by researchers from the research centre, consolidating the scientific culture of the LASIGE community.

Speakers: Jonas Bayer and Mehwish Alam
Date: June 17th, 2026, Wednesday, 12:00
Where: C6.3.27

12:00 Talk by Jonas Bayer
12:20 Talk by Mehwish Alam
12:40 Q&A + Break for snacks & coffee

Talk1: Examples of recent work in AI for mathematics
Speaker: Jonas Bayer
Summary: This talk will give an overview about my recent work in the AI, mathematics and formalisation space. I will talk about how reinforcement learning can be used to improve LLM capabilities at generating mathematical proofs in the proof assistant Lean. Moreover, I will present joint work with cognitive scientists on studying how mathematicians think when they search for proofs.
Publications: https://jonasbayer.de/#publication

Talk2: Foundational Models and Symbolic Artificial Intelligence
Speaker: Mehwish Alam
Summary: In this talk, I will give an overview of the ongoing topics which are covered in my group at Télécom Paris, Institut Polytechnique de Paris. The research spans various methods for mitigating hallucinations in Foundational Models to domain specific applications. The topics include, Chain of Though Verification, Constraint Decoding, Taxonomy Expansion, etc. with an application to the domain of biomedicine as well as cultural heritage.
Publications: https://sites.google.com/view/mehwish-alam/publications?authuser