RESEARCH PROGRAMME

One core, three applications

The Quantum Virtual Mind is the model. Everything else asks the same question of it — where is this person relative to their own baseline, and in which direction — and acts on the answer differently. Learning pushes toward the edge of capacity; restoration moves back from it; clinical research tracks departures that persist. Each card below carries its own status.

THE CORE

Quantum Virtual Mind

QVM

The QVM learns a model of the dynamics in a person’s brain–computer interface recordings — not a pattern-matcher over those recordings, but a model of the system producing them. It answers which component of the dynamics changed rather than only that something changed, and it compares a person against their own baseline rather than a population average.

What we are building

The state model, benchmarked in house against a classical baseline on public EEG data, and the agent layer that surrounds it — specialists that handle data access, model development, perturbation studies, and turning results into recommendations.

Research products

  • Cortical State ModelAn executable quantum–classical model of the dynamics a BCI can see. Physical consistency holds by construction, so it cannot produce a state the underlying system could not reach.Built & benchmarked
  • Agent LayerEight specialist agents across the lifecycle, from finding and preparing recordings through to designing what to do next. They advise; people decide.In development
  • Perturbation EngineThe shared operation underneath both applications: measure how far a state has moved from a personal baseline, and in which direction.In development

HEALTHY DIRECTION · PERFORMANCE

Superlearning

QVM → LEARNING

Adaptive learning tools watch answers, which tells you what a learner has already failed to grasp — after the fact, and without distinguishing someone who is productively stretched from someone who stopped absorbing twenty minutes ago. Reading cortical state directly addresses both gaps. Establishing that it does is our research question.

The question we are answering

Whether a state model built from a learner’s own recordings predicts readiness better than their answer history does. That study has not been run yet; designing it is the current work.

Research products

  • Superlearning Mind AtlasA map of healthy variation in the dynamics a BCI can see — between people, and within one person as they learn. A learner is a position on the map; learning is a path through it.Proposed
  • Learning Track DesignerTurns a learner’s position into a sequence: what to study next, at what difficulty, and when to stop for the day. Designed to sit behind an existing learning platform rather than replace it.Proposed
  • Cortical State LibraryA reference catalogue of distinct states across arousal, sensorimotor, cognitive and affective tiers — the vocabulary the other two products are written in.Specified

HEALTHY DIRECTION · RECOVERY

Restoration & State Regulation

QVM → WELLBEING

Guided meditation, sleep audio, and stress-regulation tools are delivered open-loop today: the same session for everyone, adjusted by self-report at best. The state a person is actually in goes unmeasured, so the content cannot respond to it. Reading arousal and affective state directly closes that loop.

The question we are answering

Whether guidance selected from a person’s measured state outperforms a fixed programme, on time to settle and on how long the settled state holds. That study has not been run.

Research products

  • Adaptive Session EngineSelects and paces guided audio from where a person’s arousal actually sits, rather than from a preset track order — slowing when they are settling, changing approach when they are not.Proposed
  • Sleep-Onset GuidanceUses the descent through arousal states that precedes sleep to time when audio should fade, instead of a fixed timer that either cuts early or plays for hours.Proposed
  • Stress Regulation TrackingMeasures whether a practice moved someone, and how far — making a wellbeing routine something you can evaluate rather than assume.Proposed

CLINICAL DIRECTION

Clinical & Mental Health Research

QVM → CLINICAL

Neurological and psychiatric conditions unfold over months and years, but the measurements taken along the way sit in unconnected documents. A state model turns that scatter into a trajectory: where a person’s dynamics sat at each visit, and how far they have moved since. This is research infrastructure for clinicians and researchers, not a diagnostic device.

The question we are answering

Whether departures from a personal baseline, measured in the model’s own components, track clinical change more sensitively than the summary scores currently used. We are seeking collaborators with longitudinal recording data.

Research products

  • Longitudinal State RecordA person’s cortical dynamics tracked across visits, so change is measured against their own history rather than a population norm.Proposed
  • Perturbation AtlasA catalogue of how specific conditions displace healthy dynamics — the clinical mirror of the learning atlas, built from the same measurements.Proposed
  • Neuromodulation PlanningFraming a candidate intervention as the inverse of a measured displacement, so a proposal can be examined before it reaches a person.Early research

HOW IT WORKS

What Superlearning actually means

Not learning faster by trying harder. Superlearning is the idea that if you can see the state a brain is in, you can stop guessing when it is ready to take on more — and stop pushing when it is not. The QVM is what makes that state visible.

Step 1 · QVM

Build the baseline

The QVM learns a model of one learner’s own dynamics from their recordings. Their baseline, not a population average.

Step 2 · Measure

Locate them today

Where do today’s dynamics sit relative to that baseline? Engaged and stretched, or past the point of absorbing anything?

Step 3 · Decide

Choose what comes next

Harder material, consolidation, or stop for the day — chosen from measured state rather than from how many questions were answered correctly.

Step 4 · Restore

Recover deliberately

Consolidation is part of learning, not a gap in it. The same state reading drives rest, sleep, and stress regulation.

Steps 1 and 2 are what the QVM does. Steps 3 and 4 are the research questions.

Mindverse Computing

The agentic stack of quantum digital twin learned from BCI data — the Quantum Virtual Mind (QVM) — an executable quantum-classical state model — wrapped in agentic workflows. An early-stage research company; everything here is research, not a medical device.

Connect

Join our newsletter

Occasional updates on our research and what we publish. No marketing.

The platform

Research directions

Company

Research and development notice: our platforms are experimental research tools for academic collaboration and proof-of-concept study. We do not provide medical diagnoses, treatment, or any regulated service. Three manuscripts are currently in preparation.

© 2026 Mindverse Computing. All Rights Reserved.