This workshop is the sixth edition of the Gradient Flows Face-to-Face series, a recurring meeting born in the aftermath of the pandemic with the aim of bringing together researchers working on gradient flows and related areas. Topics at the heart of the series include interacting particle systems, kinetic theory, variational methods for evolution equations, and optimal transport.

A distinguishing feature of this edition is a dedicated thematic focus on the theoretical analysis of transformer architectures, reflecting the rapid recent developments at the intersection of machine learning theory, optimal transport, and gradient flow methods.

As in previous editions, the workshop is designed to foster a working group atmosphere that prioritizes open discussion and close interaction among participants. The program — including the list of speakers — will be shaped collaboratively by the participants themselves, in the spirit that has defined the series since its inception.

The workshop follows five successful previous editions:
1. Rome
2. L'Aquila
3. Lyon
4. Raitenhaslach (TUM)
5. Granada