Links: arXiv Proceedings GitHub
Abstract Generative models can create entirely new images, but they can also partially modify real images in ways that …
News
All team updates, publications, blog posts and more in one place!
Academic
Generalization & Interpretability
Robust Novelty Detection through Style-Conscious Feature Ranking
Links: arXiv Proceedings GitHub
Abstract Novelty detection seeks to identify samples deviating from a known distribution, yet data shifts in a …
The decoder block of a Transformer is the basic unit of all modern LLMs. Most of the compute used for it is spent on self-attention and the MLP, with …
Academic
Generalization & Interpretability
ConceptDrift: Uncovering Biases through the Lens of Foundational Models
Links: arXiv
Abstract Datasets and pre-trained models come with intrinsic biases. Most methods rely on spotting them by analysing misclassified …
Academic
Natural Language Processing
MolMix: A Simple Yet Effective Baseline for Multimodal Molecular Representation Learning
Links: arXiv GitHub
Abstract In this work, we propose a simple transformer-based baseline for multimodal molecular representation learning, …
Academic
Generalization & Interpretability
WASP: A Weight-Space Approach to Detecting Learned Spuriousness
Links: arXiv GitHub
Abstract It is of crucial importance to train machine learning models such that they clearly understand what defines each class in …
Links: arXiv GitHub
Abstract Message-passing graph neural networks (MPNNs) have emerged as a powerful paradigm for graph-based machine learning. …
Academic
Deepfake Detection
Towards generalisable and calibrated audio deepfake detection with self-supervised representations
Links: Proceedings PDF
Abstract Generalisation—the ability of a model to perform well on unseen data—is crucial for building reliable deepfake …
Large Language Models (LLMs) took the world by storm in 2023, revolutionizing the way people search and generate text content. LLMs for code have …
Links: arXiv Proceedings GitHub
Abstract The remarkable generative capabilities of denoising diffusion models have raised new concerns regarding the …