Backrooms (A24, 2026): The Unobservable Interior of a Mind

Kane Parsons made the Backrooms mythology on YouTube before A24 gave him a feature budget and Chiwetel Ejiofor. The resulting film, released in theaters May 30, 2026, opened to a dominant weekend, reviews describing it as “a meditative voyage through the uncanny” and “Parsons’ visual representation of the human psyche disturbs like no other.”

That last phrase is not a loose metaphor. The film is, at the level of its script and visual design, a direct representation of interior mental architecture: memory degradation, behavioral loops, a space that is simultaneously infinite and claustrophobic, with no borders or tether to physical reality. The Backrooms as Parsons renders them are not an alien dimension. They are the inside of Clark’s mind made navigable.

That is not a comfortable place to be. For whole-brain emulation research, the discomfort is instructive.

The Core Shift

Clark (Chiwetel Ejiofor) is a furniture store owner operating, as the screenplay puts it, on a steady diet of rage and regret. He sees Dr. Mary Kline (Renate Reinsve) regularly to break out of repeating behavioral patterns, to forge new neural pathways, as she coaches him. Late one night, he discovers a wall in the furniture store’s basement he can pass through. On the other side is an infinite extension of the same store, a space of yellow-wallpapered rooms that never end, furnished differently in each variation, populated by distorted versions of his own memories.

The therapeutic framing at the film’s opening is precise: the goal is to recognize looping patterns and interrupt them. The Backrooms do not permit this. Every navigation choice leads back into more rooms. Every apparent exit opens onto a variation of the same interior. The Backrooms are the looping patterns made architectural.

The film’s structure is deliberately circular. Reviewer Meagan Navarro noted that “attempts to break free from circular behavioral patterns instead lead to a disturbing existential spiral.” The 1990 setting, signaled by an analog found footage cold open, positions the film outside the current moment of WBE research while keeping its core concerns precisely aligned with it.

Memory as Architecture

The Backrooms’ central conceit is that the space contains “endless possibilities, forgotten and distorted memories, and a surreal setting with no borders or tether to physical reality.” The review text identifies this explicitly as the film’s thematic operation: the mysterious space “doubles as a metaphor for the labyrinthine human mind.”

What the film depicts is not orderly memory storage. The rooms are not an archive. They are a physical instantiation of how memory actually behaves: non-indexed, non-linear, distorted by time and emotional charge, accessible only through the particular navigational logic of association rather than direct address. Clark does not search his memories the way a database query works. He wanders. He finds things by proximity, by accident, by following emotional pull rather than logical direction.

This maps onto a problem the behavioral timescale synaptic plasticity research from Nature Neuroscience 2026 identified at the cellular level. Memory encoding in the hippocampus is not the storage of discrete items in discrete locations. Single dendritic plateau potentials drive bidirectional weight changes across distributed neural populations. A memory is not an object stored in a place. It is a modification of connection weights across a network that spans millions of synapses. To capture it for whole-brain emulation, you need not just the static connectome but the current state of those synaptic weights, their precise values, and the dynamic processes by which they continue to change.

The Backrooms’ rooms shift between visits. The same room, revisited, is not identical. Objects have moved. Proportions have changed. The wallpaper pattern varies. This is cinematically precise to how hippocampal memory reconsolidation actually works: each retrieval event modifies the retrieved memory, making it slightly different from what was stored. A static snapshot of a brain’s connectome captures the weights at one moment. Any snapshot taken at a different moment captures a different configuration.

Behavioral Loops as Neural Attractors

The looping behavioral patterns the film stages are a direct analogue of attractor dynamics in neural systems. A neural attractor is a stable pattern of activity that the network tends to return to regardless of perturbations. It is the computational basis of persistent habits, recurring emotional responses, and fixed cognitive schemas: the patterns that therapy attempts to interrupt.

Clark’s loops are the film’s primary horror mechanism. He attempts to navigate differently. He chooses unexplored corridors, unfamiliar rooms. The network of rooms routes him back to familiar territory. The architecture of the Backrooms is not neutral. It has its own dynamics, and those dynamics are his dynamics. The space is not independent of the consciousness navigating it. It is constituted by that consciousness.

This distinction matters for whole-brain emulation. An emulation that captures the static connectome of a brain captures its anatomical structure, the map of which neurons connect to which others and with what approximate synaptic weights. It does not capture the dynamic attractor landscape of that connectome: the set of stable activity patterns the network falls into under different initial conditions, the particular loops that dominate under stress, the specific configurations that represent what it is like to be this person in this moment.

Erik Hoel’s 2025 formal analysis of consciousness and continual learning reached a related conclusion: consciousness may require ongoing internal adaptation as a structural feature, not as an optional addition. A static connectome snapshot cannot replicate this. The Backrooms, as navigated by Clark, are not static. They generate. Each room he enters exists because of where he came from and how he moved. The architecture is procedural, not pre-built.

The Non-Destructive Scanning Problem Rendered Cinematically

The Backrooms are specifically not observable from outside. The film’s cold open establishes that people who enter do not return in recoverable form. There is no external perspective on the interior. Dr. Kline can map Clark’s behavioral patterns from his reports and his visible behavior. She cannot observe the architecture he is navigating. The Backrooms exist in a relationship to the external world analogous to subjective experience’s relationship to third-person behavioral observation: completely opaque, accessible only from inside.

This is the physics wall problem of non-destructive brain scanning made spatial. Every high-resolution brain scanning modality currently available, fMRI, MEG, focused ultrasound, records activity at a resolution that aggregates millions of neurons per measurement point. The subjective experience produced by those neurons, the specific felt quality of what it is like to be Clark in the Backrooms, is not accessible from the measurement. The measurement records metabolic activity, local field potentials, hemodynamic responses. It does not record the Backrooms.

An emulation produced from current or near-future brain scanning would have Clark’s connectivity data and a rough measure of his regional activity patterns. It would not have the specific architecture of his Backrooms, the particular distribution of his attractor landscape, the exact geography of what he forgets and what he cannot escape. Whether the emulation would notice this absence is a different question. From the outside, it might exhibit his behavioral patterns correctly. From the inside, if there is an inside, it would be navigating a different space.

Comparative Cinematic Memory Architectures

Film	Memory model	Architecture	Accessibility	Distortion type
Backrooms (A24, 2026)	Looping attractor network	Procedurally generated, infinite	Interior only	Reconsolidation drift
Return to Silent Hill (2026)	Substrate-constituted purgatory	Fixed loop, externally indistinguishable	Interior only	Emotional stain, not factual
Necrophosis: Full Consciousness (2026)	Decaying biological vessel	Fixed but degrading	Interior + limited external	Physical substrate decay
Eternal Sunshine (2004)	Indexed archive, destructible	Spatially navigable	External technician + interior	Deliberate erasure
Inside Out (2015)	Categorized emotional memory	Organized by affect	External observer	Aging and emotional revision

The Backrooms model is the most technically accurate to contemporary memory research precisely because it is the most hostile. Memory is not a retrievable archive. It is a dynamic system that generates its contents in response to the state of the retriever. The same events produce different rooms depending on when you approach them and what emotional state you carry into the navigation.

Return to Silent Hill explores a related architecture: a substrate that constitutes the experience of the consciousness trapped within it, where the loop is not navigational but ontological. The Backrooms shares this quality. Clark cannot exit because the architecture is his architecture. There is no neutral ground to escape to.

Necrophosis: Full Consciousness takes the further step of making the degradation of the substrate visible in real time. The Backrooms degrade in a different sense: not decay but drift. The rooms shift. What was one configuration yesterday is a different configuration today. The space is alive in the sense that it changes, and it changes in response to the consciousness it contains.

What the Film Requires of WBE

The Backrooms as a framework for thinking about brain emulation produces three concrete requirements beyond structural connectome capture.

The first is dynamic state capture. A whole-brain emulation that cannot represent the current attractor landscape of its source brain cannot replicate the experience of being that brain. The connectome tells you which neurons connect to which. It does not tell you which attractor basins are deep and which are shallow, which patterns the network falls into persistently under which conditions. This requires capturing the system’s dynamic behavior, not just its static anatomy.

The second is reconsolidation dynamics. If the emulated system is going to modify its memories on retrieval, as biological memory systems do, the emulation must model not just the stored weight values but the molecular machinery by which those weights change during recall. Amyloid proteins in memory storage represent one layer of this machinery. BTSP-driven weight modification during retrieval is another. An emulation that does not replicate these dynamics will have memories that do not change on retrieval, which is not what biological memory is.

The third is the generation of interior space. The Backrooms exist because Clark’s mind generates them. They are not pre-stored. An emulation that is a static snapshot cannot generate anything. It can only replay what was captured. Whether that counts as an interior in any meaningful sense is a question the film declines to answer definitively. It does insist that the answer matters.

Limitations and Open Questions

Parsons’ film explicitly sacrifices conventional narrative resolution. The third act is widely noted as the weakest section, where screenwriter Will Soodik introduces more conventional plot mechanics that clash with the cryptic worldbuilding of the first two acts. One lead character has an unfinished arc at the film’s end. The coda teases a sequel rather than resolving the film’s philosophical questions.

This is a structural problem for analysis as well as for audiences. The film arrives at the most interesting territory, the boundary between Clark’s interior and whatever else the Backrooms might be, and stops short of exploring it. Whether the Backrooms are Clark’s mind or something that contains Clark’s mind is a distinction the film deliberately refuses to make. That ambiguity is productive philosophically but unsatisfying dramatically.

The 1990 setting removes the film from any direct dialogue with contemporary BCI research or WBE development. This is probably intentional — locating the action before the current technological trajectory keeps the film from becoming a commentary on specific technologies rather than a representation of general principles. But it means the film cannot engage directly with the specific questions current research is raising.

Ejiofor and Reinsve are consistently strong, and Parsons demonstrates significant control over composition and atmosphere. The production design is the film’s most remarkable achievement: the shifting, brightly lit yellow rooms accumulate genuine dread through repetition and variation rather than through conventional horror mechanics. Parsons is clearly building toward a larger mythology. Whether the sequel will answer the film’s deeper questions or only deepen the mystery depends on choices not yet made.

What Backrooms contributes to the WBE conversation is a spatial representation of the problem of interiority. Not the problem of scanning, or the problem of substrate, or the problem of continuity, but the prior problem: what would it mean to capture the inside of a mind, the part that cannot be observed from any external position, the architecture that is only visible to the consciousness that lives within it?

The Backrooms cannot be mapped from outside. That is not a technical limitation. It is a structural feature of what they are.

Official Sources

• Meagan Navarro. “‘Backrooms’ Review — A Disturbing Liminal Voyage Through The Human Mind.” Bloody Disgusting, May 28, 2026. URL: https://bloody-disgusting.com/reviews/3953720/backrooms-review-a-disturbing-liminal-voyage-through-the-human-mind/
• Backrooms (2026). Dir. Kane Parsons. Screenplay: Will Soodik. Stars: Chiwetel Ejiofor, Renate Reinsve. Distributor: A24. Theatrical release: May 30, 2026.
• Karl Friston. “The free-energy principle: a unified brain theory?” Nature Reviews Neuroscience, 11(2), 127-138, 2010. DOI: 10.1038/nrn2787
• Yadin Dudai. “The Neurobiology of Consolidations, Or, How Stable is the Engram?” Annual Review of Psychology, 55, 51-86, 2004. DOI: 10.1146/annurev.psych.55.090902.142050
• John Hopfield. “Neural networks and physical systems with emergent collective computational abilities.” PNAS, 79(8), 2554-2558, 1982. DOI: 10.1073/pnas.79.8.2554