A neural information element modeling and AI mapping method based on a four-dimensional discrete space-time cognitive system

Abstract

This invention relates to the fields of basic research in cognitive neuroscience, large-scale artificial intelligence model architecture, brain-computer interfaces, and educational cognitive assessment. Specifically, it relates to a unified four-dimensional discrete-time spatiotemporal modeling method for basic neural information elements of the brain (neural impulses, neuronal cluster firing, and sensory neural signals). This method is based on a homologous and unified underlying theoretical framework and is an engineering embodiment of a unified four-dimensional discrete-time spatiotemporal cognitive system. It shares a completely homologous, isomorphic, and unified mathematical framework with the inventor's previously submitted "A Token Tagging Method and Cognitive System for Input Data of a Large Language Model." The only difference is that the AI ​​patent processes computational information elements (Tokens), while this patent processes biological information elements (neural impulses, sensory signals, and neuronal firing events). Together, they constitute a unified underlying architecture covering biological cognition and artificial intelligence.

Description

Technical Field

[0001] This invention, based on the homologous and isomorphic system of previous AI patents, naturally extends to a unified brain-AI identifier architecture. This invention relates to the fields of basic research in cognitive neuroscience, large-scale artificial intelligence model architecture, brain-computer interfaces, brain rehabilitation medicine, and educational cognitive assessment. Specifically, it relates to a four-dimensional discrete spatiotemporal unified modeling method for basic neural information elements of the brain (neural impulses, neuronal cluster firing, sensory neural signals). This method is based on a homologous and unified underlying theoretical framework, simultaneously covering the needs of basic research in neuroscience neural coding and the standardized coding of information in large-scale artificial intelligence models, as well as a one-to-one precise mapping and conversion scheme between neural information elements and AI-tokens. It serves both basic research in neurophysiology and neuroscience, and provides underlying information coding support for the next generation of brain-like large-scale model architectures. It belongs to the underlying technical solution for the interoperability of brain cognition and artificial intelligence, realizing the underlying theoretical interoperability and two-way technical support between biological brain cognition and artificial intelligence information processing. Background Technology

[0002] In the field of basic research in cognitive neuroscience, the core carrier of brain cognitive activity is the individual neural impulse information element, which contains complete biological information such as firing sequence, impulse amplitude, waveform, and source pathway. It is the basic unit of neural encoding, information transmission, and cognitive formation. In essence, it is the dynamic transmission, integration, and encoding process of massive amounts of basic neural information elements (neural impulses, sensory input signals, and neuronal cluster firing events) generated by various neurons and neural pathways. Existing research has clearly identified the hippocampus-entorhinal cortex system as the core hub of spatiotemporal encoding of neural information. However, there is still a lack of a standardized, quantifiable, and globally traceable mathematical framework to achieve unified identification, temporal integration, and trajectory tracking of massive discrete neural information elements. It is difficult to systematically analyze the integration of multimodal sensory information, temporal memory encoding, cognitive state evolution, and neural signal transmission mechanisms. This results in a lack of unified analysis standards for brain science basic research data, making it difficult to systematically analyze the laws of neural information transmission. Furthermore, research data cannot be effectively integrated with artificial intelligence systems, and cross-experimental and cross-modal research results are difficult to benchmark and reuse.

[0003] In the field of large-scale artificial intelligence models, existing models generally suffer from problems such as logical inconsistencies, long-term memory loss, context collapse, and cross-device asynchrony due to the lack of token encoding with real spatiotemporal attributes and biological information. The inventors of this invention have proposed a fundamental theory for a four-dimensional discrete spatiotemporal cognitive system. By assigning unified four-dimensional discrete spatiotemporal coordinates to large-scale model tokens, this system achieves standardized encoding, global addressing, temporal control, and state synchronization of information units, fundamentally addressing the core technical deficiencies of large-scale models. However, to achieve interoperability between this AI system and brain neural information, a standardized modeling method that can accurately convert individual neural impulse information elements into AI-tokens is urgently needed.

[0004] Currently, existing technologies cannot achieve a one-to-one lossless mapping between a single neural impulse information element and an AI-token. The information encoding systems of the two major fields of brain science and artificial intelligence are completely separated, which not only restricts the quantification and engineering of brain science research, but also hinders the underlying breakthroughs in brain-like intelligence and brain-computer interface technologies. At the same time, there is a lack of a unified technical solution that takes into account both the integrity of neural information and the compatibility of AI encoding.

[0005] In summary, current basic research in brain science and the architecture of large-scale artificial intelligence models both face the technical bottleneck of lacking a unified spatiotemporal coding system for basic information units. Furthermore, the information coding frameworks of the two fields are fragmented, and existing technologies cannot achieve a one-to-one lossless mapping between individual neural impulse information elements and AI tokens. This hinders theoretical communication, data benchmarking, and technological mutual support, which restricts the in-depth analysis of neural coding mechanisms in brain science and also impedes the implementation of technologies in interdisciplinary fields such as brain-like intelligence and brain-computer interfaces. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a neural information element modeling and AI mapping method based on a four-dimensional discrete spatiotemporal cognitive system, aiming to solve the following technical problems:

[0008] To achieve full-information modeling of individual neural impulse information elements, fully preserve biological characteristics such as impulse firing timing, amplitude, waveform, and source pathway, and establish a standardized four-dimensional discrete spatiotemporal identification system to meet the quantitative analysis needs of basic brain science research;

[0009] Construct a one-to-one lossless mapping and conversion mechanism between neural information elements and AI-tokens, clarify the information capacity compatibility between the two, and realize the complete migration of neural biological information to AI computing units without losing any effective information;

[0010] It forms a common origin with the existing patent for the four-dimensional discrete spatiotemporal cognitive system of artificial intelligence, shares a unified underlying mathematical framework, and breaks down the coding barriers between brain science and artificial intelligence.

[0011] Establish a quantifiable, traceable, and reproducible neural information analysis system to support multiple applications, including neurophysiological research, cognitive state assessment, brain-like large model coding, and brain-computer interface information interaction.

[0012] Technical solution

[0013] A neural information element modeling and AI mapping method based on a four-dimensional discrete spatiotemporal cognitive system includes the following steps:

[0014] Step 1: Define the core information units and the basis for modeling.

[0015] The fundamental information unit of brain cognition is defined as a single neural impulse information element, specifically referring to an independent electrical impulse signal generated by a neuron. It fully encompasses four core biological information categories: impulse firing time, amplitude intensity, waveform characteristics, and the neural pathway to which the firing neuron belongs. No information simplification or feature compression is performed; the single impulse is the smallest modeling granularity. The meaning of a single AI-token is determined by its coordinates in a four-dimensional discrete spatiotemporal system; similarly, the meaning of a single neural impulse information element is determined by its coordinates in the same four-dimensional discrete spatiotemporal system.

[0016] The meaning of both is carried by the same type of object within the same system.

[0017] Step 2: Construct a four-dimensional discrete spatiotemporal identification system.

[0018] Adopting the same mathematical framework as the patented four-dimensional discrete spatiotemporal cognitive system for artificial intelligence, a unique four-dimensional discrete spatiotemporal coordinate (T, C, L, S) is generated or mapped for a single neural impulse information element. The parameter definitions of each dimension take into account both neurobiological attributes and AI encoding rules, achieving precise alignment between the two systems:

[0019] Time coordinate T: Identifies or maps the actual time stamp (such as Coordinated Universal Time UTC) of a single neural impulse information element, corresponding to the generation time stamp of the AI-token, serving as a unified time series reference for the dual system, and used to establish global order relationships between information units;

[0020] Context coordinate C: Identifies or maps the global cognitive context and functional stage of the neural impulse information element, defines its role attributes in the cognitive processing flow (such as input, processing, storage, output), and is isomorphic with the context temporal coordinate of AI-token, providing a unified context anchoring mechanism for the dual system;

[0021] Logical sequence coordinates L: Identify or map the relative logical order of neural impulse information elements in local neural conduction links or cognitive logical chains, corresponding to the text-internal sequence coordinates of AI-token, representing the logical relationship and structural dependency of information transmission;

[0022] Source identifier coordinates S (corresponding to device coordinates D in the original patent): identify or map the source channel and modal attributes of neural pulse information elements, and are isomorphic with the device unique identifier of AI-token, enabling accurate tracing and classification management of cross-modal and cross-source information.

[0023] Step 3: Embedding full information features of neural information elements.

[0024] The amplitude intensity, waveform characteristics, and other continuous biological features of a single neural pulse information element are embedded into the extended feature domain of the corresponding four-dimensional spatiotemporal coordinates through standardized numerical mapping, thus completely preserving all biological information of the neural pulse and forming a four-dimensional coded neural information element with full features.

[0025] Step 4: One-to-one AI-token mapping conversion.

[0026] Based on a homogeneous four-dimensional spatiotemporal identifier system, a one-to-one lossless mapping is performed between a single neural information element and a single AI-token:

[0027] The four-dimensional encoded neural information elements with full features are converted into a token format that can be recognized by the AI ​​system according to a unified mapping rule.

[0028] During the mapping process, the four-dimensional spatiotemporal coordinates, amplitude, and waveform of the neural information elements are fully preserved, with no information compression, no information loss, and no many-to-one or one-to-many merging.

[0029] The AI-token generated by the mapping is fully compatible with existing artificial intelligence four-dimensional discrete spatiotemporal cognitive systems and can be directly connected to large models to complete encoding, reasoning and memory scheduling.

[0030] Step 5: Neural information analysis and cognitive state output.

[0031] Based on the four-dimensional spatiotemporal coding results, we conduct basic research and analysis in brain science: quantify the patterns of neural impulse conduction delay, firing frequency, and amplitude changes, and analyze the neural coding mechanism; at the same time, we output cognitive state assessment results, including neural conduction efficiency, cognitive reaction speed, and memory coding strength, providing data support for brain function assessment.

[0032] Beneficial effects

[0033] 1. Precisely meets the needs of brain science research: Using a single neural pulse information element as the smallest modeling granularity, it fully preserves all biological information such as pulse amplitude, waveform, and timing, abandons information simplification processing, and constructs a quantitative modeling system that fits the essence of neurophysiology, providing a standardized analysis tool for basic research in cognitive neuroscience.

[0034] 2. Achieve homologous and isomorphic unified identification of biological information elements and AI-token: The brain and AI adopt the same four-dimensional mathematical system, and the information element addressing method is completely consistent, so that brain data can be directly mapped to the AI ​​system, constructing a complete technical chain of "neural information element modeling - AI-token mapping", and realizing the underlying theoretical communication between brain cognition and artificial intelligence.

[0035] 3. One-to-one lossless mapping with rigorous information adaptability: From the perspective of information capacity and encoding rules, a one-to-one accurate mapping between a single neural information element and a single AI-token is achieved, fully preserving all neurobiological information. This verifies that the existing AI-token capacity can fully carry the full information of the neural information element, and the technical solution has strict feasibility.

[0036] 4. Outstanding cross-domain application value: It can be directly used for quantitative analysis of EEG physiological data and assessment of neurological function, and can also provide a token encoding scheme with biological spatiotemporal attributes for the next generation of brain-like large models. It can also support bidirectional information interaction of brain-computer interfaces, taking into account both basic research and engineering implementation value. Attached Figure Description

[0037] Appendix Figure 1 This is a flowchart illustrating the overall process of neural information element modeling and AIToken mapping as described in an embodiment of the present invention.

[0038] Appendix Figure 2 This is a schematic diagram illustrating the one-to-one lossless mapping principle between neural information elements and AIToken as described in an embodiment of the present invention;

[0039] Appendix Figure 3 This is a block diagram of the neural information modeling and AIToken mapping system described in an embodiment of the present invention. Detailed Implementation

[0040] This embodiment provides a neural information element modeling and AI mapping method based on a four-dimensional discrete spatiotemporal cognitive system. This embodiment is only used to explain the present invention and is not intended to limit the scope of protection of the present invention.

[0041] The acquisition of neural pulse information elements involves using neurophysiological acquisition equipment to obtain single neural pulse signals generated by brain neurons. A single neural discharge is defined as an independent neural pulse information element, and its corresponding amplitude, waveform, and other biophysical characteristics are recorded.

[0042] The four-dimensional discrete spatiotemporal coordinate assignment is a unique four-dimensional discrete spatiotemporal coordinate (T, C, L, S) assigned to the aforementioned neural impulse information element, where:

[0043] T represents the actual timestamp of the neural pulse information element, using UTC unified timing.

[0044] C represents the cognitive context and functional processing stage of the information element, used to distinguish the processing state of the signal;

[0045] L is the relative logical sequence number of the information element in the local neural conduction link;

[0046] S represents the source neural pathway and acquisition channel identifier corresponding to this information element.

[0047] Biometric embedding standardizes the continuous features of neural pulse information elements, such as amplitude intensity and waveform width, and embeds them into the extended feature domain corresponding to the four-dimensional coordinates to form a neural information coding unit containing complete information.

[0048] The one-to-one mapping with AI-Token establishes a one-to-one lossless mapping relationship between the neural information elements carrying four-dimensional coordinates and complete biological characteristics and a single token in the artificial intelligence system. This maintains the same origin and structure of the coordinate system, without information aggregation, discarding, or compression, and achieves unified alignment between brain signals and AI representations.

[0049] Cognitive spatiotemporal trajectory construction is based on the four-dimensional coordinates of multiple neural information elements to construct their evolutionary trajectory in the spatiotemporal dimension, which can be used for cognitive state analysis, neural pathway tracking, or brain-computer interface interaction.

Claims

1. Claim 1 (Independent Claim): A method for mapping neural information meta-identifiers and AI tokens based on a four-dimensional discrete spatiotemporal system, characterized in that... include: Acquire neural pulse information elements, with a single neural pulse as the smallest information granularity; Assign a unique four-dimensional discrete spatiotemporal coordinate (T, C, L, S) to the neural impulse information element; The biological characteristics of the neural impulses are embedded into the feature domain corresponding to the four-dimensional coordinates; Perform a one-to-one lossless mapping between the neural information elements and the AI ​​Token; The cognitive spatiotemporal trajectory is constructed based on four-dimensional coordinates and the analysis output is completed.

2. Claim 2 (dependent claim): The method according to claim 1, characterized in that, The four-dimensional discrete spatiotemporal coordinates (T, C, L, S) are of the same origin and structure as the four-dimensional discrete spatiotemporal cognitive system of artificial intelligence, covering all coordinate combinations and expression forms under this system, and realizing unified addressing of brain signals and AI representations.

3. Claim 3 (dependent claim) The method according to claim 1, characterized in that, The one-to-one lossless mapping is a bidirectional mapping, which includes converting neural information elements into AI tokens and reversing the AI ​​tokens into neural information element format signals.

4. Claim 4 (dependent claim): The method according to claim 1, characterized in that, The source identifier coordinate S corresponds to the unique device identifier in the artificial intelligence system, enabling accurate tracing of information across modalities and sources.

5. Claim 5 (dependent claim) The method according to claim 1, characterized in that, The method is applicable to various neurophysiological signal acquisition scenarios and all types of AI model architectures, and does not rely on specific algorithms.

6. Claim 6 (Independent Claim): A neural information and AI token mapping system, characterized in that, include: The system includes a neural signal acquisition module, a four-dimensional coordinate encoding module, a biometric embedding module, a one-to-one mapping and transformation module, a spatiotemporal trajectory analysis module, and an AI system interface. Each module works together to execute the method described in any one of claims 1 to 5.

7. Claim 7 (Independent Claim): A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method described in any one of claims 1 to 5.

Images