From generative models to generative passages: A computational approach to (neuro)phenomenology
PsyArXiv February 23, 2021 preprint DOI: 10.31234/osf.io/k9pbn
Summary
This paper proposes a computational approach to neurophenomenology by using generative models, particularly large language models, to produce phenomenological descriptions of lived experience. The authors argue that such models can generate rich, structured passages that bridge first-person subjective reports and third-person neural data, offering a new method for studying consciousness. They illustrate this with examples and discuss implications for integrating computational techniques with phenomenological inquiry.
Study at a glance
| Design | theoretical or philosophical paper |
|---|---|
| Key finding | Generative models can be used to produce phenomenological passages that may help integrate first-person and third-person perspectives in the study of consciousness. |
Abstract
From generative models to generative passages: A computational approach to (neuro)phenomenology