1 citations · 1 across the 11 of their papers we have counts for
8 papers · 1 filter
Weight Oracles: Reading Neural Network Weights with Language Models
Krishna Kabra, Constantin Venhoff, Christian Schroeder de Witt
Interpretability methods for neural networks are predominantly reactive: they analyse activations produced during specific forward passes, requiring known inputs to find hidden cap…
A Low-Rank Subspace Analysis of LLM Interventions
Angira Sharma, Christian Schroeder de Witt, Philip Torr +2
Interventions designed to modify a particular behavior in LLMs, such as refusal or sycophancy, often produce unintended changes in other behaviors. This lack of targeted control ma…
When Language Representations Interact: Separability and Cross-Lingual Effects in LLMs
Boris Marinov, Angira Sharma, Christian Schroeder de Witt +3
Large language models exhibit strong multilingual capabilities, however, their internal representations are difficult to interpret. Understanding these interactions is important fo…
Sensitivity Analysis for Climate Science with Generative Flow Models
Alex Dobra, Jakiw Pidstrigach, Tim Reichelt +6
Sensitivity analysis is a cornerstone of climate science, essential for understanding phenomena ranging from storm intensity to long-term climate feedbacks. However, computing thes…
DEEDEE: Fast and Scalable Out-of-Distribution Dynamics Detection
Tala Aljaafari, Varun Kanade, Philip Torr +1
Deploying reinforcement learning (RL) in safety-critical settings is constrained by brittleness under distribution shift. We study out-of-distribution (OOD) detection for RL time s…
Mixture of Experts Made Intrinsically Interpretable
Xingyi Yang, Constantin Venhoff, Ashkan Khakzar +4
Neurons in large language models often exhibit \emph{polysemanticity}, simultaneously encoding multiple unrelated concepts and obscuring interpretability. Instead of relying on pos…
PSyDUCK: Training-Free Steganography for Latent Diffusion
Aqib Mahfuz, Georgia Channing, Mark van der Wilk +3
Recent advances in generative AI have opened promising avenues for steganography, which can securely protect sensitive information for individuals operating in hostile environments…
SAGE: Scalable Ground Truth Evaluations for Large Sparse Autoencoders
Constantin Venhoff, Anisoara Calinescu, Philip Torr +1
A key challenge in interpretability is to decompose model activations into meaningful features. Sparse autoencoders (SAEs) have emerged as a promising tool for this task. However,…