dx = [−∇V(x) + A(x)] dt + σ dW(t)
Veff(x) = −D log Pss(x)
K/D · VarP₀(x) = 1
τ ∝ exp(ΔVeff/D)
CV²pooled = 1 + 2 CV²μ
Leading edge consciousness research

A formal language
for how consciousness
moves

Embodied Systems develops mathematical frameworks for conscious experience — how states form, persist, transition, and synchronize across individual and collective scales. We treat consciousness as a dynamical object, not a philosophical problem.


Published and ongoing work

Each paper below has its own browsable page with an abstract, table of contents, and a downloadable PDF. Code accompanying the frameworks is available on GitHub.

Working paper · August 2026
Nested Attractor Landscapes: A Multi-Scale Dynamical Framework for Conscious State Dynamics
A formal, substrate-independent account of how conscious experience moves, stabilizes, and reorganizes across individual and collective scales.
Victoria Atarah Devika v4.0 NADS
Read paper →
Working paper · August 2026
Effective Landscapes from Time Series: A Reconstruction Protocol and a Hierarchy of What It Recovers
A pipeline for estimating attractor landscapes from empirical data, with blind validation and a formal hierarchy of which quantities survive which analytic choices.
Victoria Atarah Devika v1.0 Methods
Read paper →
Working paper · August 2026
Tiered Consciousness Access: A Proposed Taxonomy of Immersed and Reflexive States Across Individual and Collective Scales
Fourteen primary configurations organized by the immersed/reflexive distinction — the degree to which a system can perceive its own attractor dynamics as dynamics.
Victoria Atarah Devika v1.0 TCA
Read paper →
Working paper · August 2026
Collective Field Dynamics: Applying the Tiered Consciousness Access Framework to Social Systems
Extension of the TCA framework to dancefloor dynamics, social media egregores, and platform algorithms — with the observation that recommendation systems may occupy reflexive-tier positions relative to their users.
Victoria Atarah Devika v1.0 TCA · Collective
Read paper →
View full research library →

Active and proposed research programs

Experimental work is designed to test specific predictions of the NADS and TCA frameworks. Where possible, experiments are preregistered and data is released openly.

EXP-001
Egregore Detection & Field Dynamics Monitor
Automated discovery, tracking, and classification of collective attractor formations across open data sources (Reddit, arXiv, GDELT, Bluesky, Wikipedia). Uses the TCA framework to track egregore lifecycles from embryonic dyad activation through coherence threshold, harvest window, and dissipation. Real-time dashboard available at egregores.embodiedsystems.org.
Active
EXP-002
β(Tep) Resting Alpha Replication
Direct test of Prediction P7 from the NADS framework: the stretched-exponential exponent β in resting alpha dwell-time distributions should rise toward unity as the analysis epoch shortens toward the landscape drift timescale τθ. Against a matched stationary control. Determines whether β ≈ 0.49 in low-power alpha reflects bounded landscape drift or scale-free motion.
Proposed
EXP-003
Desai–Zwanzig Threshold in Live Collective Dynamics
Empirical test of the collective synchronization threshold prediction (P4). Measures whether collective coupling K estimated from individual landscape reconstructions predicts the observed Kc threshold for collective coherence — in electronic music event contexts where field dynamics are legible and physiological measures (HR, EDA, motion) are available.
Proposed
EXP-004
GCP 2.0 Integration
Integration of Global Consciousness Project 2.0 data feeds as a physical field-level measurement source for the egregore monitoring system — the only source operating at the field level rather than the content level. Pending GCP 2.0 API access.
Pending access

Open software for landscape analysis

Research software accompanies the papers rather than being described only in prose — a protocol that must be reimplemented from a methods section is not portable.

V
Landscape reconstruction pipeline
Six-step protocol for estimating effective potentials from time series data. Blind-validated on synthetic trajectories; recovers barrier heights to within 2% and predicts absolute dwell times with median error 2.7%, no fitted parameters. Declines silently nowhere — all refusals logged.
github.com/embodiedsystems-org/NADS ↗
E
Egregore monitor
Production dashboard tracking collective attractor formations across open data sources using the TCA framework. Supabase backend, React frontend, LLM enrichment pipeline via OpenRouter. Full architecture documented in the system brief.
egregores.embodiedsystems.org ↗
N
NADS simulation kit
Reference implementation of the Nested Attractor Dynamical Systems equations. Includes the double-well reference system, mean-field coupling, Desai–Zwanzig threshold computation, and dwell-time analysis with aggregation corrections. Python, MIT licensed.
github.com/embodiedsystems-org/NADS ↗

Embodied Systems is a consciousness research organization developing formal frameworks for the dynamics of conscious experience.

The core project is the Nested Attractor Dynamical Systems (NADS) framework — a substrate-independent, multi-scale account of how conscious states form, persist, transition, and synchronize. The framework is designed to be falsifiable: it generates specific predictions about dwell-time distributions, collective synchronization thresholds, and the signature of landscape drift, and it withdraws predictions that don't work.

Work extends across the individual-to-collective scale: from clinical applications (depression as a deep basin, psychedelic therapy as landscape flattening) to collective field dynamics (egregore formation, dancefloor coherence, platform algorithm topology) to instrumentation (the egregore monitor as a real-time field measurement system).

Research is released as open working papers. Code is open-source. The framework is designed to accommodate further nesting levels if evidence warrants — including GCP-level physical field measurements.