← Search

Cian Eastwood

8 accepted papers

2025

Towards scientific discovery with dictionary learning: Extracting biological concepts from microscopy foundation models

ICML 2025poster

Sparse dictionary learning (DL) has emerged as a powerful approach to extract semantically meaningful concepts from the internals of large language models (LLMs) trained mainly in the text domain. In this work, we explore whether DL can extract meaningful concepts from less human-interpretable scien…

Cited by 1SourcePDFScholar
2023

DCI-ES: An Extended Disentanglement Framework with Connections to Identifiability

ICLR 2023poster

In representation learning, a common approach is to seek representations which disentangle the underlying factors of variation. Eastwood & Williams (2018) proposed three metrics for quantifying the quality of such disentangled representations: disentanglement (D), completeness (C) and informativenes…

2023

Spuriosity Didn’t Kill the Classifier: Using Invariant Predictions to Harness Spurious Features

NeurIPS 2023poster

To avoid failures on out-of-distribution data, recent works have sought to extract features that have an invariant or stable relationship with the label across domains, discarding "spurious" or unstable features whose relationship with the label changes across domains. However, unstable features oft…

Cited by 19SourcePDFScholar
2022

Probable Domain Generalization via Quantile Risk Minimization

NeurIPS 2022accept

Domain generalization (DG) seeks predictors which perform well on unseen test distributions by leveraging data drawn from multiple related training distributions or domains. To achieve this, DG is commonly formulated as an average- or worst-case problem over the set of possible domains. However, pre…

Cited by 76SourcePDFScholar
2022

Source-Free Adaptation to Measurement Shift via Bottom-Up Feature Restoration

ICLR 2022spotlight

Source-free domain adaptation (SFDA) aims to adapt a model trained on labelled data in a source domain to unlabelled data in a target domain without access to the source-domain data during adaptation. Existing methods for SFDA leverage entropy-minimization techniques which: (i) apply only to classif…

2020

Learning Object-Centric Representations of Multi-Object Scenes from Multiple Views

NeurIPS 2020spotlight

Learning object-centric representations of multi-object scenes is a promising approach towards machine intelligence, facilitating high-level reasoning and control from visual sensory data. However, current approaches for \textit{unsupervised object-centric scene representation} are incapable of aggr…

2018

A Framework for the Quantitative Evaluation of Disentangled Representations

ICLR 2018poster

Recent AI research has emphasised the importance of learning disentangled representations of the explanatory factors behind data. Despite the growing interest in models which can learn such representations, visual inspection remains the standard evaluation metric. While various desiderata have been…