Implementing surrogate goals for safer bargaining in LLM-based agents

TL;DR

This paper explores implementing surrogate goals in language models to promote safer bargaining behavior, comparing prompting, fine-tuning, and scaffolding methods through experiments.

Contribution

It introduces and evaluates four methods for implementing surrogate goals in language models, highlighting scaffolding and fine-tuning as most effective.

Findings

Scaffolding and fine-tuning outperform prompting in implementing surrogate goals.

Fine-tuning and scaffolding more accurately reflect desired threat responses.

Scaffolding-based methods have the best overall performance.

Abstract

Surrogate goals have been proposed as a strategy for reducing risks from bargaining failures. A surrogate goal is goal that a principal can give an AI agent and that deflects any threats against the agent away from what the principal cares about. For example, one might make one's agent care about preventing money from being burned. Then in bargaining interactions, other agents can threaten to burn their money instead of threatening to spending money to hurt the principal. Importantly, the agent has to care equally about preventing money from being burned as it cares about money being spent to hurt the principal. In this paper, we implement surrogate goals in language-model-based agents. In particular, we try to get a language-model-based agent to react to threats of burning money in the same way it would react to "normal" threats. We propose four different methods, using techniques of prompting, fine-tuning, and scaffolding. We evaluate the four methods experimentally. We find that methods based on scaffolding and fine-tuning outperform simple prompting. In particular, fine-tuning and scaffolding more precisely implement the desired behavior w.r.t. threats against the surrogate goal. We also compare the different methods in terms of their side effects on capabilities and propensities in other situations. We find that scaffolding-based methods perform best.