My favourite papers

Societal Adaptation to Advanced AI [pdf]
Jamie Bernardi, et al. ArXiv, 2024

The Statistical Thermodynamics of Generative Diffusion Models [pdf]
Luca Ambrogioni. ArXiv, 2023

Should we Automate Democracy? [pdf]
Johannes Himmelreich. The Oxford Handbook of Digital Ethics, 2022

Attention Approximates Sparse Distributed Memory [pdf]
Trenton Bricken & Cengiz Pehlevan. NeurIPS, 2021

The Vulnerable World Hypothesis [pdf]
Nick Bostrom. Global Policy, 2019

Dissecting Adam: The Sign, Magnitude, and Variance of Stochastic Grads [pdf]
Lukas Balles & Philipp Hennig. ICML, 2018

Structured Access: An Emerging Paradigm for Safe AI Deployment [pdf]
Toby Shevlane. ArXiv, 2022