Seminars

The alignment of large language models (LLMs) can often be brittle when faced with the complexities of real-world deployment. In this talk, I share our investigations on two scenarios where special care is required to ensure robust alignment.

"Join via MS Teams":https://teams.microsoft.com/l/meetup-join/19%3ameeting_ZDE0NzFlMmYtODgyNi00N2M5LThhZGMtZGY2Zjk3M2FiOGJm%40thread.v2/0?context=%7b%22Tid%22%3a%2249a50445-bdfa-4b79-ade3-547b4f3986e9%22%2c%22Oid%22%3a%22c74ff4ca-98fe-4b28-9889-e119acc12f30%22%7d

The seminar will be an interdisciplinary journey between applications of Artificial Intelligence (AI) Computer Vision based methods to the field of Forensics and Biomedical Imaging.

End-to-end encryption (E2EE) for messaging has become an industry standard and is widely implemented in many applications. However, applying E2EE to audio calls, particularly group calls, remains a complex challenge.

Equivariance can enhance the data efficiency of machine learning models by incorporating prior knowledge about a problem.
Thanks to their flexibility and generality, steerable CNNs are a popular design choice for equivariant networks.

*Abstract*

If you have a question about this talk, please contact Anand Rao Tadipatri.

=== Hybrid talk ===

Join Zoom Meeting https://cam-ac-uk.zoom.us/j/89856091954?pwd=Bba77QB2KuTideTlH6PjAmbXLO8HbY.1

Meeting ID: 898 5609 1954 Passcode: ITPtalk

Abstract not available

We present Aeneas, a verification toolchain for Rust programs based on a
lightweight functional translation.
We leverage Rust's rich region-based type system to generate pure models for a large class
of safe Rust programs which can contain shared and mutable borrows, functions returning

I will introduce the All-atom Diffusion Transformer (ADiT), a unified generative modelling architecture capable of jointly modelling both periodic crystals and non-periodic molecular systems.