The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets

NeurIPS 2023 Workshop ATTRIB Submission45 Authors

Published: 27 Oct 2023, Last Modified: 08 Dec 2023ATTRIB PosterEveryoneRevisionsBibTeX
Keywords: Large Language Models, interpretability, knowledge representations, probing, causal interventions, truthfulness
TL;DR: Investigates the way LLMs represent the difference between true and false statements, and improves our ability to extract LLM beliefs.
Abstract: Large Language Models (LLMs) have impressive capabilities, but are also prone to outputting falsehoods. Recent work has developed techniques for inferring whether a LLM is telling the truth by training probes on the LLM's internal activations. However, this line of work is controversial, with some authors pointing out failures of these probes to generalize in basic ways, among other conceptual issues. In this work, we curate high-quality datasets of true/false statements and use them to study in detail the structure of LLM representations of truth, drawing on three lines of evidence: 1. Visualizations of LLM true/false statement representations, which reveal clear linear structure. 2. Transfer experiments in which probes trained on one dataset generalize to different datasets. 3. Causal evidence obtained by surgicallly intervening in a LLM's forward pass, causing it to treat false statements as true and vice versa. Overall, we present evidence that language models linearly represent the truth or falsehood of factual statements. We also introduce a novel technique, mass-mean probing, which generalizes better and is more causally implicated in model outputs than other probing techniques.
Submission Number: 45