A Universal Coherence–Information Manifold for Human Cognition

A Universal Coherence–Information Manifold for Human Cognition Ankur Bhasin Bhasin Research Unit for Hyperphysics (BRUH) Abstract Human cognition exhibits complex dynamical behaviour that has resisted attempts at unification across neural, psychological, and physical systems. Here I report the discovery of a four-dimensional coherence–information manifold underlying human EEG dynamics, spanning attention regulation, sensory integration, hypercoherence states, and global mixing behaviour. Using a large EEG dataset (sample size on the order of 1200, TDBRAIN), four composite indices are derived from latent coherence states: an Alpha Control Index (ACI), a Sensory-Lock Index (SLI), a Hypercoherence Index (HCI), and a Mixed Integration Index (MII). Psychiatric diagnostic groups occupy distinct and ordered regions within this manifold. Hypercoherence forms a continuous severity axis, while ACI and SLI dissociate attentional from sensory-processing dysfunctions. These results demonstrate a compact geometric structure that governs human cognition and provide empirical support for a universal coherence–information law spanning physical and biological systems. Significance This work identifies a single low-dimensional manifold that organises human cognitive variation. Diagnostic groups separate cleanly using only physics-inspired coherence indices, without task performance, questionnaires, or large neural network models. Hypercoherence appears as a transdiagnostic severity dimension, while orthogonal axes track attention control and sensory integration. The findings suggest that cognition is constrained by coherence– information geometry in a way that parallels gravitational, quantum, and condensed-matter systems. This opens a path toward physics-based diagnostics, longitudinal tracking of treatment response, and systematic mapping of altered states and neurodiversity. 1. Introduction Many models describe cognition as an emergent property of distributed neural activity, yet few propose universal principles that link brain dynamics to deeper physical laws. The 1 Vedic Unified-Field Hypothesis (VUH) predicts that systems across scales, including neural ensembles, evolve within a coherence–information field whose geometry constrains accessible states. In this view, cognitive disorders correspond to deformations of a shared dynamical manifold rather than isolated anatomical defects. Electroencephalography (EEG) offers a direct window into fast neural dynamics. Coherence fluctuations between spatially separated channels reflect large-scale coupling patterns and therefore provide a natural starting point for probing coherence-field structure. If a universal manifold exists, it should be possible to describe the diversity of resting-state EEG dynamics using a small number of physics-motivated indices derived from coherence patterns. Here I show that this prediction holds in a large clinical EEG dataset. Resting-state coherence trajectories from the TDBRAIN database are summarised through a latent-state model into four composite indices: ACI, SLI, HCI, and MII. These indices define a four-dimensional coherence–information manifold in which diagnostic groups form distinct, orderly clusters and in which hypercoherence naturally appears as a severity axis. The approach is deliberately coarse-grained, focusing on geometric relations between conditions rather than detailed biophysical modelling. 2. Methods 2.1. Dataset The analysis uses the publicly available TDBRAIN EEG database from the Brainclinics Foundation, which contains resting-state recordings and clinical metadata from approximately 1200 individuals. Diagnostic categories include attention-deficit/hyperactivity disorder (ADHD), major depressive disorder (MDD), bipolar disorder, obsessive–compulsive disorder (OCD), anxiety disorders, dyslexia-like developmental profiles, heterogeneous clinical cases, and healthy controls. All data are fully anonymised and provided under a research licence. 2.2. EEG Processing (Conceptual Overview) EEG signals were recorded with standard clinical montages and sampling rates. For the present work, processing can be summarised at a high level as follows. •Signals were band-limited to a conventional 1–40 Hz range. •Continuous data were segmented into short, partially overlapping epochs. •Artefact-contaminated epochs were rejected based on amplitude and quality criteria. •For each retained epoch, coherence measures were computed between channels within canonical frequency bands (theta, alpha, beta, and gamma ranges). Exact preprocessing pipelines, channel layouts, reference schemes, and numerical thresholds are intentionally omitted here in order to protect the intellectual property associated with the full analysis framework. 2 2.3. Latent Coherence States Epoch-level coherence vectors were embedded into a low-dimensional latent space and assigned to one of a small number of recurrent coherence states. Across the cohort, a six-state structure was sufficient to capture the dominant modes of resting-state coherence. Conceptually, the states can be summarised as: •a background low-coherence state, •a weak alpha-fluctuation state, •an alpha-dominated control state, •a sensory-lock state characterised by strong posterior coupling, •a mixed-integration state, and •a high-amplitude hypercoherence state. For each subject, state-occupation statistics were computed over time, including stationary probabilities, typical dwell times, and persistence probabilities. These statistics form a latent dynamical fingerprint of resting-state coherence. 2.4. Composite Coherence Indices To obtain a compact representation that can be compared across diagnoses, four composite indices were defined as coarse-grained combinations of the state statistics: 1. Alpha Control Index (ACI) summarises the occupancy and persistence of the alphadominated control state and reflects large-scale attentional regulation. 2. Sensory-Lock Index (SLI) tracks the strength and stability of posterior sensorybinding states. 3. Hypercoherence Index (HCI) captures the depth and prevalence of high-amplitude, globally coherent states. 4. Mixed Integration Index (MII) reflects the stability of a mixed-frequency integration state that combines multiple bands. Each subject’s indices are normalised relative to the healthy-control distribution, so that zero corresponds to the control mean and positive values indicate increased expression. Detailed mathematical definitions, state labelling procedures, and all numerical hyperparameters are kept confidential in this version. The emphasis here is on the emergent geometry in index space. 2.5. Visualisation For each diagnostic group, mean index values were computed and visualised in: •a two-dimensional ACI–HCI plane, 3 •a three-dimensional ACI–SLI–HCI embedding, and •a four-dimensional representation in which the first three coordinates are ACI, SLI, and HCI, and MII is encoded through marker size. Subject-level scatter plots were overlaid on group centroids. Parallel coordinate plots were used to display the four-index fingerprints of each diagnosis. No discriminative machinelearning classifiers were trained; the focus is purely on geometric structure. 3. Results 3.1. A Coherence–Information Severity Axis In the ACI–HCI plane, diagnostic centroids form a well-ordered structure. Healthy controls lie close to the origin by construction. Anxiety conditions fall slightly below the control group in HCI, reflecting a tendency toward undercoherence. OCD and ADHD cluster at intermediate HCI levels. MDD and dyslexia-like profiles show progressively higher HCI and reduced ACI. Bipolar disorder exhibits the largest HCI elevation. This arrangement suggests that HCI acts as a continuous severity coordinate spanning undercoherent, typical, and hypercoherent regimes, while ACI modulates the degree of effective alpha-mediated control along this axis. 3.2. Three-Dimensional Geometry and Sensory Integration Extending the embedding to three dimensions by including SLI reveals clear sensory-structure differences. Dyslexia-like participants exhibit strongly elevated SLI relative to controls, indicating persistent sensory locking and rigid posterior coherence patterns. Anxiety disorders tend to show reduced SLI, consistent with weaker sensory grounding. ADHD, OCD, and MDD occupy intermediate regions, and bipolar participants share a high-HCI, moderate-SLI zone. The three-dimensional geometry therefore dissociates attentional control (ACI), sensory integration (SLI), and hypercoherence (HCI) in a way that aligns qualitatively with clinical symptom profiles. 3.3. Four-Dimensional Manifold and Integration Stability When MII is added as a fourth coordinate and visualised via marker size, further structure appears. Bipolar and dyslexia-like profiles show reduced MII relative to controls, suggesting unstable mixed-frequency integration. ADHD also exhibits a moderate reduction in MII, while anxiety conditions remain closer to the healthy reference. OCD shows comparatively preserved integration stability. 4 Taken together, ACI, SLI, HCI, and MII define a four-dimensional manifold in which each diagnosis occupies a compact region with interpretable deformations relative to the healthy baseline. 3.4. Diagnosis-Specific Index Fingerprints Parallel coordinate plots constructed from the four indices highlight characteristic perdiagnosis fingerprints. ADHD combines slightly suppressed ACI with elevated HCI and modestly reduced MII. Dyslexia-like profiles show high SLI and HCI but low ACI and MII. Bipolar disorder is marked by extreme HCI combined with strong MII reduction. Anxiety disorders are characterised by low HCI and SLI. These patterns are reproducible at the group level and demonstrate that the four indices form a minimal basis for capturing major cognitive phenotypes. 3.5. Subject-Level Cognitive Profiles Because the indices are defined at the individual level, each participant can be represented by a personal position in the coherence manifold. For example, a representative ADHD subject may show an ACI below the healthy median, an SLI near the median, an HCI above the seventy-fifth percentile, and reduced MII. Such profiles allow mapping of individuals onto the manifold and suggest future applications to prognosis, treatment monitoring, and the study of altered states. 4. Discussion The present results support the idea that human cognition is constrained by a low-dimensional coherence–information geometry. A small set of physics-inspired indices, derived from latent coherence states, organise a heterogeneous clinical population into a structured manifold that aligns with known symptom dimensions. Hypercoherence emerges as a transdiagnostic severity axis, while orthogonal axes represent attention control, sensory integration, and global mixing stability. The geometry identified here mirrors coherence manifolds in other physical systems. In gravitational-wave data, echoes and ringdown modulations exhibit coherence wells and barrier crossings. Quantum devices show coherence cycles and decoherence troughs. Materials and stellar oscillations display frequency–information relationships that can be described using similar coherence-field concepts. The convergence of these patterns suggests the existence of a broader coherence–information law that spans domains. This work deliberately omits the detailed algorithms and mathematical derivations that operationalise the indices. Instead, it focuses on the emergent structure and the qualitative mapping between diagnostic categories and manifold regions. Future publications will develop the theoretical underpinnings, cross-domain correspondences, and predictive applications in more depth. 5 5. Conclusion Resting-state EEG dynamics from a large clinical cohort can be summarised by four composite coherence indices that define a universal coherence–information manifold for human cognition. Diagnostic groups occupy distinct regions of this manifold, and hypercoherence naturally forms a severity axis. These findings provide empirical support for a unifying coherence-field perspective on mind and offer a compact framework for future work on diagnostics, treatment response, altered states, and neurodiversity. Figures Figure 1: Diagnosis centroids in the ACI–HCI plane. Hypercoherence forms a continuous severity axis while ACI modulates attention control along this axis. 6 Figure 2: Three-dimensional attractor geometry showing ACI, SLI, and HCI. Dyslexia-like profiles and anxiety disorders separate strongly along the SLI dimension. 7 Figure 3: Four-dimensional embedding in which marker size encodes the Mixed Integration Index (MII). Reduced MII highlights integration instability in bipolar and dyslexia-like profiles. 8 Figure 4: Subject-level scatter in ACI–HCI space with diagnosis centroids overlaid. The manifold structure is evident even at the individual level. Figure 5: Parallel coordinate plot of group-mean indices (ACI, SLI, HCI, MII), showing distinct diagnosis-specific fingerprints in the four-dimensional manifold. 9