Evaluating Theory of (an uncertain) Mind: Predicting the Uncertain Beliefs of Others in Conversation Forecasting

TL;DR

This paper introduces new tasks for language models to predict and quantify the uncertainty of others' beliefs in dialogue, emphasizing the complexity of modeling uncertain mental states.

Contribution

It presents a novel suite of conversation forecasting tasks that challenge LMs to model uncertain beliefs, incorporating rescaling, variance reduction, and demographic context.

Findings

LMs explain up to 7% variance in uncertainty prediction

Tasks are challenging, indicating room for improvement

Experiments conducted on three diverse dialogue corpora

Abstract

Typically, when evaluating Theory of Mind, we consider the beliefs of others to be binary: held or not held. But what if someone is unsure about their own beliefs? How can we quantify this uncertainty? We propose a new suite of tasks, challenging language models (LMs) to model the uncertainty of others in dialogue. We design these tasks around conversation forecasting, wherein an agent forecasts an unobserved outcome to a conversation. Uniquely, we view interlocutors themselves as forecasters, asking an LM to predict the uncertainty of the interlocutors (a probability). We experiment with re-scaling methods, variance reduction strategies, and demographic context, for this regression task, conducting experiments on three dialogue corpora (social, negotiation, task-oriented) with eight LMs. While LMs can explain up to 7% variance in the uncertainty of others, we highlight the difficulty of the tasks and room for future work, especially in practical applications, like anticipating ``false