BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Learning to Discover Learning Algorithms - Alex Goldie - Universit
 y of Oxford 
DTSTART:20260309T130000Z
DTEND:20260309T140000Z
UID:TALK245179@talks.cam.ac.uk
CONTACT:Sally Matthews
DESCRIPTION:Abstract: Automating the development of machine learning algor
 ithms (i.e.\, meta-learning) has the potential to unlock new frontiers in 
 the field. However\, our ability to learn to discover has been limited by 
 a focus on small\, static benchmarks. Motivated by how procedural generati
 on unlocked generalist agents in reinforcement learning\, this talk will e
 xplore how a similar approach can be applied to algorithm discovery in mac
 hine learning. Specifically\, I will introduce DiscoGen\, a new procedural
  generator of algorithm discovery tasks. Using DiscoGen\, we demonstrate h
 ow agents used for algorithm discovery can themselves be optimised in a me
 ta-meta-loop. DiscoGen further establishes principled task design for the 
 field\, in particular emphasising the need for meta-train and meta-test di
 stinctions. Finally\, the talk will discuss future research ideas enabled 
 by DiscoGen\, such as training algorithm world models or other means for o
 ptimising discovery agents.\n\n\nBio: Alexander D. Goldie is a final-year 
 PhD student at the University of Oxford\, co-supervised by Jakob N. Foerst
 er (FLAIR) and Shimon Whiteson (WhiRL). His research focuses on the automa
 ted discovery of machine learning algorithms (meta-learning)\, using black
 -box evolution\, symbolic evolution\, or agentic LLMs. More recently\, he 
 has explored how agents can themselves be optimised in a meta-meta-learnin
 g process.
LOCATION:Computer Lab\, SS03
END:VEVENT
END:VCALENDAR