We are very pleased to announce Bernardino Romera-Paredes (Google DeepMind) as our second keynoter for this year’s edition of PPSN! He is a former core team member of AlphaFold2 and AlphaTensor and now research scientist at Google DeepMind in London.

At PPSN 2024 Bernardino Romera-Paredes will present his current research regarding the evolution of new heuristics, supported by pairing a pre-trained LLM and an automated “evaluator”, within his keynote:

FunSearch: Discovering new mathematics and algorithms using Large Language Models

In this talk I will present FunSearch, a method to search for new solutions in mathematics and computer science. FunSearch works by pairing a pre-trained LLM, whose goal is to provide creative solutions in the form of computer code, with an automated “evaluator”, which guards against hallucinations and incorrect ideas. By leveraging these two components within an evolutionary algorithm, initial solutions “evolve” into new knowledge. I will present the application of FunSearch to a central problem in extremal combinatorics — the cap set problem — where we discover new constructions of large cap sets going beyond the best known ones, both in finite dimensional and asymptotic cases. This represents the first discoveries made for established open problems using LLMs. Then, I will present the application of FunSearch to an algorithmic problem, online bin packing, which showcases the generality of the method. In this use case, FunSearch finds new heuristics that improve upon widely used baselines. I will conclude the talk by discussing the implications of searching in the space of code.

For more information about Bernardino Romera-Paredes go to https://www.romera-paredes.com/.