Francis Rhys Ward

The Elic­i­ta­tion Game: Eval­u­at­ing ca­pa­bil­ity elic­i­ta­tion techniques

Why care about AI per­son­hood?

[Paper] AI Sand­bag­ging: Lan­guage Models can Strate­gi­cally Un­der­perform on Evaluations

An In­tro­duc­tion to AI Sandbagging

Sim­ple dis­tri­bu­tion ap­prox­i­ma­tion: When sam­pled 100 times, can lan­guage mod­els yield 80% A and 20% B?

Tall Tales at Differ­ent Scales: Eval­u­at­ing Scal­ing Trends For De­cep­tion In Lan­guage Models

Re­ward Hack­ing from a Causal Perspective

Agency from a causal perspective

Causal­ity: A Brief Introduction

In­tro­duc­tion to Towards Causal Foun­da­tions of Safe AGI

For ev­ery choice of AGI difficulty, con­di­tion­ing on grad­ual take-off im­plies shorter timelines.

On Agent In­cen­tives to Ma­nipu­late Hu­man Feed­back in Multi-Agent Re­ward Learn­ing Scenarios