This post summarizes work done over the summer as part of the Summer 2023 AI Safety Hub Labs programme. Our results will also be published as part of an upcoming paper. In this post, we focus on explaining how we define and evaluate properties of deceptive behavior in LMs and present evaluation results for state-of-the-art models. The work presented in this post was done by the AISHL group of Felix Hofstätter, Harriet Wood, Louis Thomson, Oliver Jaffe, and Patrik Bartak under the supervision of Francis Rhys Ward and with help from Sam Brown (not an AISHL participant).

1 Introduction

For a long time, the alignment community has discussed the possibility that advanced AI systems may learn to deceive humans. Recently, the evidence has grown that language models (LMs) deployed in real life may be capable of deception. GPT-4 pretended to be a visually impaired person to convince a human to solve a CAPTCHA. Meta's Diplomacy AI Cicero told another player it was on the phone with its girlfriend after a server outage. Many large models are sycophantic, adapting their responses based on a user's preferences, even if that means being less truthful. Recent work shows that LMs lie and can be prompted to act deceptively. To make sure that advanced AI systems are safe, ideally, we would like to be able to evaluate if they are deceptive. Our work proposes purely behavioral methods for evaluating properties related to deception in language models. We use them to demonstrate scaling trends for these properties.

The examples above intuitively feel like deception, but it is unclear if they would fit a more rigorous definition. In Section 2, we will explain the operationalization of deception that we are using and discuss related work. This formalization requires assessing the agency, beliefs, and intentions of LMs. We argue that consistency of beliefs is an important property of agency. We also briefly argue that the LMs we evaluate act with intent but leave a proper evaluation of LM intent to future work. In Section 3, we define belief behaviorally and propose different methods for eliciting a model's beliefs. Building on this definition, we operationalize lying in LMs. In Section 4, we present the results of applying our belief evaluations to state-of-the-art LMs. We show that when more compute is spent on either training or inference, LMs...

Posts

Wikitag Contributions

Comments