Can "consciousness" be observed from large language model (LLM) internal states? Dissecting LLM representations obtained from Theory of Mind test with Integrated Information Theory and Span Representation analysis
arXiv Preprint Archive June 26, 2025 preprint
Study at a glance
AI-extracted from the abstract| Characteristics | Systematic analysis |
|---|---|
| Population | Sequences of Large Language Model (LLM) representations from Theory of Mind test results |
| Keywords | Cs.cl Cs.ai Cs.ne Q-bio.nc Ai consciousness Artificial consciousness Machine consciousness Ai sentience Ai awareness Large language models Llms Generative ai Neural networks Deep learning Natural language processing Nlp Integrated information theory Iit Theories of consciousness Consciousness research Information representation Internal states Structural patterns Neural representations |
| Key points | Sequences of contemporary Transformer-based LLM representations lack statistically significant indicators of observed consciousness phenomena. |
Abstract
Integrated Information Theory (IIT) provides a quantitative framework for explaining consciousness phenomenon, positing that conscious systems comprise elements integrated through causal properties. We apply IIT 3.0 and 4.0 -- the latest iterations of this framework -- to sequences of Large Language Model (LLM) representations, analyzing data derived from existing Theory of Mind (ToM) test results. Our study systematically investigates whether the differences of ToM test performances, when presented in the LLM representations, can be revealed by IIT estimates, i.e., $\Phi^{\max}$ (IIT 3.0), $\Phi$ (IIT 4.0), Conceptual Information (IIT 3.0), and $\Phi$-structure (IIT 4.0). Furthermore, we compare these metrics with the Span Representations independent of any estimate for consciousness. This additional effort aims to differentiate between potential "consciousness" phenomena and inherent separations within LLM representational space. We conduct comprehensive experiments examining variations across LLM transformer layers and linguistic spans from stimuli. Our results suggest that sequences of contemporary Transformer-based LLM representations lack statistically significant indicators of observed "consciousness" phenomena but exhibit intriguing patterns under $\textit{spatio}$-permutational analyses. The Appendix and code are available as Supplementary Materials at: https://doi.org/10.1016/j.nlp.2025.100163.