Limits to Predicting Online Speech Using Large Language Models

TL;DR

This study investigates the limits of predicting online speech with large language models, revealing that user-specific context greatly enhances prediction accuracy and that certain social features are learned in-context, with consistent results across diverse demographics.

Contribution

The paper provides a comprehensive analysis of predictability limits in modeling user-generated content on Twitter, highlighting the importance of user history and social features in language model performance.

Findings

Predicting individual user posts remains challenging.

Using user history significantly improves prediction accuracy.

Models learn social features like @-mentions and hashtags in-context.

Abstract

Our paper studies the predictability of online speech -- that is, how well language models learn to model the distribution of user generated content on X (previously Twitter). We define predictability as a measure of the model's uncertainty, i.e. its negative log-likelihood. As the basis of our study, we collect 10M tweets for ``tweet-tuning'' base models and a further 6.25M posts from more than five thousand X (previously Twitter) users and their peers. In our study involving more than 5000 subjects, we find that predicting posts of individual users remains surprisingly hard. Moreover, it matters greatly what context is used: models using the users' own history significantly outperform models using posts from their social circle. We validate these results across four large language models ranging in size from 1.5 billion to 70 billion parameters. Moreover, our results replicate if instead of prompting the model with additional context, we finetune on it. We follow up with a detailed investigation on what is learned in-context and a demographic analysis. Up to 20\% of what is learned in-context is the use of @-mentions and hashtags. Our main results hold across the demographic groups we studied.