Research question
How does a person experience vivid images, environments, people, sounds, and events during dreams while the physical eyes are closed and not observing those environments?
How the project is being studied
The research tracks individual claims against their sources, the evidence offered, competing interpretations, and remaining uncertainty. It follows a sequence: define the question, establish the conventional neuroscience account, trace sources, challenge each claim, and only then form a synthesis.
Dream reports, neurophysiology, memory experiments, and computational models answer different questions. The project keeps those forms of evidence separate.
Current evidence boundary
Dreaming is studied through reports of subjective experience and measurements of sleep-related brain activity. How those measures relate to the content of a specific dream remains an active question.
Some computational accounts propose that internally generated dream-like experiences could help organize learned representations. A model's behavior is not direct proof that human dreams use the same mechanism.
The working connection between waking prediction errors, memory updating, and later dream experience has not been established as a single causal mechanism. In particular, this project has not found direct evidence that prediction error inside dreams causes memory reconsolidation.
Questions being tested next
- Which brain processes support internally generated visual and multisensory experience during sleep?
- How do REM and non-REM dream experiences differ, and what evidence distinguishes them?
- What experiments, if any, directly connect prediction error, later dream content, and measurable memory updating?
Sources under review
These readings frame the questions; listing them does not mean every proposal in them has been confirmed.
- Nir & Tononi, “Dreaming and the brain: from phenomenology to neurophysiology” (2010) — review of dream experience and neurophysiology.
- Deperrois et al., “Learning beyond sensations: How dreams organize neuronal representations” (2024) — review and theoretical proposals.
- Deperrois et al., “Learning cortical representations through perturbed and adversarial dreaming” (2022) — computational model.