2 citations · 2 across the 4 of their papers we have counts for
4 papers
Corrigibility Transformation: Constructing Goals That Accept Updates
Rubi Hudson
An AI agent will learn a desired goal more effectively if it does not resist the training process, but many partially learned goals incentivize an AI to avoid further goal updates.…
Joint Scoring Rules: Zero-Sum Competition Avoids Performative Prediction
Rubi Hudson
In a decision-making scenario, a principal could use conditional predictions from an expert agent to inform their choice. However, this approach would introduce a fundamental confl…
Incentivizing honest performative predictions with proper scoring rules
Caspar Oesterheld, Johannes Treutlein, Emery Cooper +1
Proper scoring rules incentivize experts to accurately report beliefs, assuming predictions cannot influence outcomes. We relax this assumption and investigate incentives when pred…
Conditioning Predictive Models: Risks and Strategies
Evan Hubinger, Adam Jermyn, Johannes Treutlein +2
Our intention is to provide a definitive reference on what it would take to safely make use of generative/predictive models in the absence of a solution to the Eliciting Latent Kno…