Mindscape Collective is now The Consciousness Library. Same library, new name. You may need to sign in again. About the change
Skip to content

Adriana Michalak

4 papers in the library · 1 citation · publishing 2023-2026

Papers

Automatic Scoring of Dream Reports' Emotional Content with Large Language Models

arXiv (Cornell University) February 28, 2023 Lorenzo Bertolini, Valentina Elce, Adriana Michalak et al. 1 citation

Dream research usually depends on human experts manually scoring dream reports, a time-consuming process. While natural language processing tools have been explored for automatic analysis, they could not reason over a report's full context, needed extensive preprocessing, and were rarely validated against manual scoring. This work used large language models, both off-the-shelf and bespoke, to replicate manual annotation of dream reports, focusing on emotions. The off-the-shelf method performed poorly, likely due to linguistic differences across individuals. In contrast, the bespoke text classification method achieved high performance and was robust against biases. This approach may enable analysis of large dream datasets and improve reproducibility and comparability across studies.

Immersive NREM dreaming preserves subjective sleep depth against declining sleep pressure

bioRxiv (Cold Spring Harbor Laboratory) September 8, 2025 Adriana Michalak, Davide Marzoli, Francesco Pietrogiacomi et al. preprint

Perceived sleep depth, a key factor in subjective sleep quality, is traditionally linked to unconsciousness and reduced brain activity. Using high-density EEG and a serial awakening paradigm during NREM sleep, researchers found that deeper sleep corresponded to lower high-to-low frequency power ratio, indicating reduced cortical activation. However, this relationship weakened when dreaming occurred, suggesting immersive conscious experiences can counteract the effect of cortical activation on perceived depth. Perceived sleep depth was lowest during states with a mere sense of presence and highest during immersive dreaming or deep unconsciousness. As the night progressed, physiological sleep pressure and subjective sleepiness declined, but perceived sleep depth increased alongside rising dream immersiveness. These findings challenge the view that deep sleep stems solely from reduced brain activity.

Immersive NREM2 dreaming preserves subjective sleep depth against declining sleep pressure.

PLoS Biology March 1, 2026 Adriana Michalak, Davide Marzoli, Francesco Pietrogiacomi et al.

Perceived sleep depth, a key part of subjective sleep quality, is traditionally linked to unconsciousness and reduced brain activity. Combining high-density EEG with a serial awakening paradigm during N2 sleep in healthy participants, the study found that deeper sleep was associated with a lower high-to-low frequency power ratio, indicating reduced cortical activation. However, this relationship weakened when dreaming occurred, suggesting immersive conscious experiences can counteract the impact of cortical activation on perceived depth. Perceived sleep depth was lowest during minimal awareness (a mere sense of presence) and highest during immersive dreaming or deep unconsciousness.

Automatic Annotation of Dream Report’s Emotional Content with Large Language Models

Workshop on Computational Linguistics and Clinical Psychology January 1, 2024 Lorenzo Bertolini, V. Elce, Adriana Michalak et al.

Dream reports are typically analyzed by trained human annotators, a time-consuming process. While earlier natural language processing tools could automate some analysis, they could not reason over full report context, required extensive preprocessing, and were rarely validated against manual scoring. This study used large language models (LLMs) to replicate manual annotation of dream reports, focusing on references to emotions. An off-the-shelf LLM method performed poorly, likely due to linguistic differences between reports from different individuals. A bespoke text classification method achieved high performance and was robust against biases. The approach may enable analysis of large dream datasets and improve reproducibility across studies.