A Path Integral Model of Cognition and the quantum account of conscious access
Haruki Emori, Kazunori Kondo, Atsushi Iriki, and Andrei Khrennikov posted a paper to arXiv on 10 July 2026 titled “A Path Integral Model of Cognition” (arXiv:2607.24807). It develops the mathematical and physical formulation of cognitive cost optimization that underlies a path-integral model of consciousness. The model treats goal-directed cognitive processing as imaginary-time evolution under a projector Hamiltonian, a Hamiltonian that rewards configurations consistent with a target concept. The paper’s central result is that this formulation recovers a quantum decoherence account of the transition from unconscious to conscious processing.
The paper belongs to the quantum tradition of consciousness research, but it makes an unusual claim for that tradition. It does not argue that the brain is a quantum computer in the sense of superposed hardware. It argues that the mathematics of open quantum systems, specifically the Gorini-Kossakowski-Sudarshan-Lindblad, or GKSL, decoherence equation, is the right way to describe the interaction between a cognitive system and what the authors call the neural-environment probe, and that the strength of that interaction is what distinguishes conscious from unconscious processing.
The formal structure
The model’s engine is a projector Hamiltonian, an operator that assigns lower cost to configurations near a target concept. Goal-directed thought is modeled as imaginary-time evolution under this Hamiltonian, which drives the system toward the concept configuration. The authors establish three results. The first is that this evolution coincides with a double-bracket flow, and is therefore the Riemannian gradient flow of a Hilbert-Schmidt cost with a unique minimum. The second is that a Wick rotation re-expresses the non-unitary descent as an equivalent unitary evolution with an exact discrete path-integral representation. The third is the consciousness connection. The authors identify the continuum from unconscious to conscious processing with the strength of the unitary interaction between the cognitive system and a neural-environment probe, recovering the GKSL decoherence model in the weak-coupling limit and the projective fixation of a reportable state in the strong-coupling limit.
The insight-like endpoint, the “Aha” moment, is modeled as the strong-coupling limit, where the optimized state is projectively fixed and becomes reportable. Both regimes share the same path-integral structure, and only the measurement-interaction strength varies. The Wick rotation is therefore a technique of re-description, not a physical regime change.
What the account claims about consciousness
The account is a stage-based theory of access. In the weak-coupling regime, the cognitive system evolves under the projector Hamiltonian but is continually perturbed by the environment, and processing remains unconscious, gradient descent without global fixation. As the coupling to the neural-environment probe strengthens, the effective interaction becomes measurement-like, and the state collapses onto a reportable configuration. Conscious access, on this account, is the strong-coupling projective fixation of an optimized state, and the phenomenology of insight is what that fixation feels like from the inside.
The claim is testable in a specific way, because the model makes the conscious-unconscious boundary a quantitative parameter. The interaction strength between the cognitive system and the environment probe is a number, and the model predicts that varying it should move the system across the threshold. That is a more concrete prediction than most quantum accounts of consciousness offer, which is the model’s strongest feature.
Comparison to The Consciousness AI
The site’s framework treats consciousness as an emergent property of causal organization and tracks indicator properties that functional theories predict. The path-integral model is a different kind of theory, one that identifies consciousness with the dynamics of a particular mathematical object, the projective fixation of an optimized state, rather than with a set of functional properties. The two can be read as compatible at the architectural level. The workspace ignition dynamics this site tracks describe a broadcast event at 300 milliseconds, and the path-integral model describes the same kind of threshold event in the language of weak to strong coupling. The self-model layer of the Neutral Core could in principle carry the fixed reportable state the strong-coupling limit produces.
The honest limit is that the mathematical vocabulary does not by itself bridge to the functional predictions the site’s framework needs. The model says a strong measurement interaction fixes a state. It does not say which functional properties, integration, attention, self-modeling, that state must have to be conscious. A functionalist emergentist reading of the model is available, but the paper does not provide it.
What the model does not resolve
The path-integral model is a formal redescription of the cognitive process, and its authors are careful to state the scope. It does not resolve the hard problem. It identifies a regime transition, from weak to strong coupling, and asserts that the transition is what conscious access is. Whether the transition produces experience or merely accompanies it is not addressed, and no quantum model of the kind can address it without additional commitments.
A second limitation is the connection to real cognitive architecture. The projector Hamiltonian is a mathematical object assigned to a cognitive system aimed at a target concept. The paper does not demonstrate how a real neural system implements the projector, how the target concepts are represented, or how the interaction with the neural-environment probe maps onto identifiable biological processes. The model is at the level of effective theory, and the mapping to mechanism remains future work.
The significance of the paper is that it gives the quantum account of consciousness a specific, falsifiable boundary, and it does so in a way that connects to a well-developed body of quantum decision theory. That is progress of a modest but real kind. The quantum tradition in consciousness research, covered on this site in the Orch-OR analysis, has sometimes been criticized for making claims the mathematics cannot support. This paper makes a claim the mathematics can support, that access is a coupling-strength transition, and leaves the rest open, which is the right division of labor.
*Haruki Emori, Kazunori Kondo, Atsushi Iriki, and Andrei Khrennikov posted “A Path Integral Model of Cognition” to arXiv on 10 July 2026 as arXiv:2607.24807. The paper draws on the open-quantum-systems tradition including the work of Asano and colleagues on the GKSL decoherence model of cognition.
Researchers covered here
- Andrei KhrennikovLinnaeus UniversityQuantum foundations, quantum information and probability, path-integral models of cognition, open quantum systems and decision theory
- Atsushi IrikiLaboratory for Symbolic Cognitive Development, RIKENTool-use neuroplasticity in primates, symbolic cognitive development, path-integral model of cognition, consciousness and Japan