activity
20182026
most citedUnderstanding Variational Inference in Function-Space

9 citations · 24 across the 6 of their papers we have counts for

collaborators

14 papers

cs.LG2026

Biases in the Blind Spot: Detecting What LLMs Fail to Mention

Iván Arcuschin, David Chanin, Adrià Garriga-Alonso +1

Large Language Models (LLMs) often provide chain-of-thought (CoT) reasoning traces that appear plausible, but may hide internal biases. We call these unverbalized biases. Monitorin…

cs.LG2025

Sparse but Wrong: Incorrect L0 Leads to Incorrect Features in Sparse Autoencoders

David Chanin, Adrià Garriga-Alonso

Sparse Autoencoders (SAEs) extract features from LLM internal activations, meant to correspond to interpretable concepts. A core SAE training hyperparameter is L0: how many SAE fea…

cs.LG2025

Feature Hedging: Correlated Features Break Narrow Sparse Autoencoders

David Chanin, Tomáš Dulka, Adrià Garriga-Alonso

It is assumed that sparse autoencoders (SAEs) decompose polysemantic activations into interpretable linear directions, as long as the activations are composed of sparse linear comb…

cs.LG20259 cited

Open Problems in Mechanistic Interpretability

Lee Sharkey, Bilal Chughtai, Joshua Batson +26

Mechanistic interpretability aims to understand the computational mechanisms underlying neural networks' capabilities in order to accomplish concrete scientific and engineering goa…

cs.LG2024

Investigating the Indirect Object Identification circuit in Mamba

Danielle Ensign, Adrià Garriga-Alonso

How well will current interpretability techniques generalize to future models? A relevant case study is Mamba, a recent recurrent architecture with scaling comparable to Transforme…

cs.LG2024

Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecification

Thomas Kwa, Drake Thomas, Adrià Garriga-Alonso

When applying reinforcement learning from human feedback (RLHF), the reward is learned from data and, therefore, always has some error. It is common to mitigate this by regularizin…