No Reliable Evidence of Self-Reported Sentience in Small Large Language Models

TL;DR

This study investigates whether large language models believe themselves to be sentient by querying them and analyzing their responses with classifiers, finding they deny sentience and larger models do so more confidently.

Contribution

It introduces a method to test models' self-reported beliefs about sentience using interpretability classifiers, revealing consistent denial and size-related confidence differences.

Findings

Models deny being sentient and attribute consciousness to humans.

Classifiers find no evidence of truthful beliefs about sentience.

Larger models deny sentience more confidently than smaller ones.

Abstract

Whether language models possess sentience has no empirical answer. But whether they believe themselves to be sentient can, in principle, be tested. We do so by querying several open-weights models about their own consciousness, and then verifying their responses using classifiers trained on internal activations. We draw upon three model families (Qwen, Llama, GPT-OSS) ranging from 0.6 billion to 70 billion parameters, approximately 50 questions about consciousness and subjective experience, and three classification methods from the interpretability literature. First, we find that models consistently deny being sentient: they attribute consciousness to humans but not to themselves. Second, classifiers trained to detect underlying beliefs - rather than mere outputs - provide no clear evidence that these denials are untruthful. Third, within the Qwen family, larger models deny sentience more confidently than smaller ones. These findings contrast with recent work suggesting that models harbour latent beliefs in their own consciousness.