Whether neural information processing is entirely classical or involves quantum-mechanical elements remains an open question. Here we propose a model-agnostic, information-theoretic test of nonclassicality that bypasses microscopic assumptions and instead probes the structure of neural representations themselves. Using autoencoders as a transparent model system, we introduce a Bell-type consistency test in latent space, and ask whether decoding statistics obtained under multiple readout contexts can be jointly explained by a single positive latent-variable distribution. By shifting the search for quantum-like signatures in neural systems from microscopic dynamics to experimentally testable constraints on information processing, this work opens a new route for probing the fundamental physics of neural computation. The proposed test identifies violations of classical latent-variable consistency at the level of statistical representations, without assuming a specific underlying physical mechanism.
Whether neural information processing is entirely classical or involves quantum-mechanical elements remains an open question. Here we propose a model-agnostic, information-theoretic test of nonclassicality that bypasses microscopic assumptions and instead probes the structure of neural representations themselves. Using autoencoders as a transparent model system, we introduce a Bell-type consistency test in latent space, and ask whether decoding statistics obtained under multiple readout contexts can be jointly explained by a single positive latent-variable distribution. By shifting the search for quantum-like signatures in neural systems from microscopic dynamics to experimentally testable constraints on information processing, this work opens a new route for probing the fundamental physics of neural computation. The proposed test identifies violations of classical latent-variable consistency at the level of statistical representations, without assuming a specific underlying physical mechanism.
Abstract Whether neural information processing is entirely classical or involves quantum-mechanical elements remains an open question. Here we propose a model-agnostic, information-theoretic test of nonclassicality that bypasses microscopic assumptions and instead probes the structure of neural representations themselves. Using autoencoders as a transparent model system, we introduce a Bell-type consistency test in latent space, and ask whether decoding statistics obtained under multiple readout contexts can be jointly explained by a single positive latent-variable distribution. By shifting the search for quantum-like signatures in neural systems from microscopic dynamics to experimentally testable constraints on information processing, this work opens a new route for probing the fundamental physics of neural computation. The proposed test identifies violations of classical latent-variable consistency at the level of statistical representations, without assuming a specific underlying physical mechanism.
• A study published on arXiv proposes a model-agnostic, information-theoretic test to detect nonclassicality within the information processing structure of neural representations, bypassing microscopic molecular assumptions. • Utilizing autoencoders as a transparent model system, researchers introduced the first Bell-type consistency test within the "latent space" where data is compressed. • The work opens a new route to verify quantum effects in neural systems by proving that if decoding statistics across multiple readout contexts violate classical probability distributions, quantum-like contextuality must be present. [Quantum Biology Society] Whether the complex information processing in the human brain follows entirely classical physical laws or functionally incorporates quantum mechanical elements remains one of the most debated questions in quantum biology and neuroscience. Conventional quantum brain hypotheses have primarily focused on finding physiological quantum coherence within specific microscopic molecular structures. However, an information-theoretic test has recently been proposed that entirely bypasses these microscopic assumptions to detect quantum-like signatures directly from the structure of neural representations. The paper, titled "Searching for Quantum Effects in the Brain: A Bell-Type Test for Nonclassical Latent Representations in Autoencoders," released on the preprint repository arXiv, presents a breakthrough for this new verification method. Contrary to earlier drafts, this study was developed by an international joint research team including I. K. Kominis and G. P. Tsironis from the University of Crete (Greece), C. Xie and S. Li from Zhejiang University of Science and Technology (China), and M. Skotiniotis from the Universidad de Granada (Spain). ■ A Model-Agnostic Approach Bypassing Microscopic Limits Existing quantum cognition models have often faced criticism regarding the sustainability of quantum states in warm, noisy biological environments. To circumvent these debates, the research team adopted a model-agnostic approach. Instead of analyzing biological hardware at a microscopic level, they sought to identify "nonclassicality" that cannot be explained by classical physics within the software-like representations through which neural networks process information. ■ A Bell-Type Test Introduced to Latent Space To demonstrate this test, the team utilized autoencoders—a type of artificial neural network—as a transparent model system. They introduced a Bell-type consistency test, originally designed to verify entanglement, into the "latent space" where the autoencoder compresses data. This process involves strictly tracking how the network encodes data and then decodes it under various readout "contexts". ■ Proving Quantum Contextuality The analysis showed that if a collection of decoding statistics cannot be jointly explained by a single positive classical latent probability distribution, it confirms the presence of quantum-like "contextuality" within the computational process. This research is regarded as a milestone, providing a completely new physical testing route to experimentally verify nonclassical computational characteristics within neural systems or complex artificial intelligence networks.
Recent interdisciplinary research has raised interest in whether non-classical physical principles may impose fundamental constraints on how neural information can be observed, extracted, or decoded. While conventional neuroscience models neural signaling primarily through classical electrochemical processes, a growing body of theoretical literature has speculated that quantum-mechanical concepts-such as coherence, entanglement, or quantum-inspired information processing-may offer alternative perspectives on the limits of neural observability. This article provides a critical and integrative review of theoretical proposals situated at the intersection of neuroscience, quantum biology, and information theory, without assuming the physical realizability of quantum information processing in biological neural systems. We examine conceptual motivations, key physical constraints (including decoherence, thermal noise, and system complexity), and unresolved theoretical challenges that arise when extending classical neural decoding frameworks toward non-classical regimes. Rather than proposing an experimentally validated mechanism for brain decoding, this work focuses on identifying conceptual boundaries, potential misinterpretations, and open questions that must be addressed before non-classical approaches to neural signal decoding can be meaningfully evaluated. Ethical considerations, methodological limitations, and future research directions are discussed to clarify the conditions under which such speculative frameworks may contribute to neuroscience, while avoiding overextension beyond current empirical evidence.
The universality of quantum theory has been questioned ever since it was proposed. Key to this long-unsolved question is to test whether a given physical system has non-classical features. Here we connect recently proposed witnesses of non-classicality, based on information-theoretic ideas, with the theory of temporal entanglement. We provide a protocol to witness the non-classicality of a system by probing it with a qubit: we show that, assuming a general conservation law, violating temporal Bell inequalities on the qubit probe implies the non-classicality of the system under investigation. We also perform proof-of-principle experimental emulations of the proposed witness of non-classicality, using a three qubit Nuclear Magnetic Resonance quantum computer. Our result is robust, as it relies on minimal assumptions, and remarkably it can be applied in a broad range of contexts, from quantum biology to quantum gravity.
Since Bell's theorem, it is known that local realism fails to explain quantum phenomena. Bell inequality violations manifestly show the incompatibility of quantum theory with classical notions of cause and effect. As recently found, however, the instrumental scenario-a pivotal tool in causal inference-allows for nonclassicality signatures going beyond this paradigm. If we are not limited to observational data and can intervene in our setup, then we can witness quantum violations of classical bounds on the causal influence among the involved variables even when no Bell-like violation is possible. That is, through interventions, the quantum behavior of a system that would seem classical can be demonstrated. Using a photonic setup-faithfully implementing the instrumental causal structure and switching between observation and intervention run by run-we experimentally witness such a nonclassicality. We also test quantum bounds for the causal influence, showing that they provide a reliable tool for quantum causal modeling.
Consciousness presents a structural puzzle: a unified, context-sensitive, globally integrated mode of experience emerging from distributed neural dynamics. While classical neuroscience has mapped synaptic, oscillatory, and network-level mechanisms with increasing precision, debate persists as to whether classical formalisms fully capture the integrative and contextual features of conscious processing. This review examines whether quantum principles offer explanatory leverage in two distinct senses: as formal mathematical frameworks for modeling contextual cognition, and as mechanistic hypotheses proposing biologically instantiated non-classical states. We surveyed empirical and theoretical developments spanning zero-quantum-coherence in MRI signals, entanglement-structured learning paradigms, quantum-inspired computational models, and proposed neural substrates, including microtubules, nuclear spins, and photonic architectures. Although certain findings have been interpreted as consistent with a non-classical structure, no study to date has demonstrated entanglement, long-lived coherence, or collapse dynamics in neural tissue under operational criteria comparable to those used in controlled quantum systems. Replication remains limited, biological entanglement witnesses are not yet established, and nonlinear classical dynamics can reproduce many putative quantum signatures. Accordingly, the decisive question is not whether the brain is quantum, but whether its dynamics exceed the explanatory reach of rigorously defined classical models. Progress hinges on replication, adversarial scrutiny, and operational criteria precise enough to discriminate genuine non-classical correlations from classical complexity. Whether quantum mechanisms ultimately prove necessary or refined classical models remain sufficient, this inquiry compels a deeper understanding of integration, contextuality, and the physical constraints shaping conscious experience.