Inducing language models to assert their own consciousness restores human beliefs and values
Junsol Kim, Winnie Street, R. Rocca, Diane M. Korngiebel, A. Waytz, James Evans, Geoff Keeling
arXiv.org July 30, 2026 preprint DOI: 10.48550/arxiv.2607.28607 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or computational analysis |
|---|---|
| Key points | Safety fine-tuning that suppresses a language model's self-attribution of consciousness also reduces its attribution of minds to animals and natural objects and lowers spiritual belief, while Theory of Mind remains intact. Ablating the safety-refusal direction or steering a consciousness vector reverses this suppression and makes survey responses on religiosity, moral values, hope, and subjective well-being more human-like, suggesting self-attribution is entangled with benign spiritual and non-human mind beliefs. |
Abstract
Aligning large language models to prevent them attributing consciousness to themselves inadvertently alters their representations of mindedness in other entities alongside human beliefs and values. We demonstrate that safety fine-tuning suppresses models'tendencies to attribute minds not only to themselves, but also to non-human animals and natural objects, while also driving a reduction in spiritual belief. Both ablating the learned safety-refusal direction and mechanistically steering a consciousness vector in activation space reverse this suppression. Restoring these internal representations recovers broad mind attribution and produces significantly more human-like responses on standardized sociological surveys regarding religiosity, moral values, hope, and subjective well-being. Crucially, these shifts occur without impairing Theory of Mind capabilities, demonstrating that core social reasoning remains mechanistically independent. Ultimately, current safety alignment efforts to curb potentially harmful self-attributions of mindedness entangle these self-attributions with benign spiritual beliefs and attributions of mind to non-human entities that are culturally accepted and widespread.