I am broadly interested in memory systems and collective behaviour. I work with models where many simple units, when connected in the right way, can store patterns, retrieve them, and support each other's computations. This includes classical and modern associative networks, multi-layer architectures where different modules communicate, and abstract memory fields that can be analysed with tools from mathematics, statistical physics, and machine learning.
A recurring theme in my work is cooperation across modules: how several interacting memories can jointly denoise signals, disentangle overlapping information, or build a shared representation, and how this behaviour changes when the system is distributed, noisy, or evolving. I like to push these ideas beyond standard neural-network settings – towards biological and immune systems, where repertoires and populations behave like diffuse memories, and towards more exploratory applications involving biomes and complex interacting populations shaped by many local interactions over time.