6 papers
Stress Testing Concept Erasure with Large Language Model Agents
Yuyang Xue, Feng Chen, Zhihua Liu +4
Concept erasure aims to remove semantic concepts from a trained generative model and is increasingly important for responsible AI deployment. However, verifying whether a model has…
Optimal L2 Regularization in High-dimensional Continual Linear Regression
Gilad Karpel, Edward Moroshko, Ran Levinstein +3
We study generalization in an overparameterized continual linear regression setting, where a model is trained with L2 (isotropic) regularization across a sequence of tasks. We deri…
A Causal Framework for Mitigating Data Shifts in Healthcare
Kurt Butler, Stephanie Riley, Damian Machlanski +13
Developing predictive models that perform reliably across diverse patient populations and heterogeneous environments is a core aim of medical research. However, generalization is o…
A model predictive control framework with customer-priority tiers for virtual power plant resilience during extreme weather: A UK heatwave case study
Edward Moroshko, Weizhe Qin, Desen Kirli +3
Due to changes in frequency and intensity of extreme weather events, such as heatwaves and storms, power systems around the globe are having to deal with increased imbalance betwee…
A Shift in Perspective on Causality in Domain Generalization
Damian Machlanski, Stephanie Riley, Edward Moroshko +7
The promise that causal modelling can lead to robust AI generalization has been challenged in recent work on domain generalization (DG) benchmarks. We revisit the claims of the cau…
CRCE: Coreference-Retention Concept Erasure in Text-to-Image Diffusion Models
Yuyang Xue, Edward Moroshko, Feng Chen +3
Text-to-Image diffusion models can produce undesirable content that necessitates concept erasure. However, existing methods struggle with under-erasure, leaving residual traces of…