JP7779509B2 - Gel processing device and gel processing method - Google Patents

Description

The present invention relates to a self-learning computer and a self-learning method.

For a century, particle physicists and astrophysicists have argued that self-operating nature describes its information as geometric shapes hidden in a time structure. Inspired by the idea of cellular automatons, Feynman wrote in 1965 that the laws of physics would one day be replaced by changing geometric shapes (Non-Patent Document 1). There is another historic challenge of the century: the idea of learning unknown data simply by mapping periodic events dates back several decades.

R. P. Feynman, The Character of Physical Law (Modern Library). (MIT Press, Cambridge, MA, 1965), pp. 173.

There have been a century of rigorous theoretical and experimental efforts to map various phenomena (e.g., the brain's conscious experience, earthquakes, weather forecasts, deep-space astrophysical events, market shares, and viral information) using structures of periodic events or loops known as time crystals. None of these historical efforts have been truly successful.

One aspect of the present invention was made in consideration of the above problems, and one of its objectives is to provide an information processing device for self-learning that can memorize various phenomena using a periodic structure or loop.

To achieve the above-mentioned objectives, an information processing device for self-learning includes a container, a gelatinous material placed within the container, and a plurality of antennas arranged around the container to generate an electromagnetic field corresponding to input data.

To achieve the above-mentioned objective, an information processing method for self-learning includes the steps of preparing a gel-like material to be placed in a container, generating an electromagnetic field corresponding to input data, and supplying the electromagnetic field to multiple antennas arranged around the container.

In accordance with one aspect of the present invention, periodic events can be filtered out as helical nanowires of a specific length, pitch, and diameter. Helical nanowires can support a variety of phenomena using periodic structures, i.e., loops.

FIG. 1 is a schematic diagram illustrating a step-by-step approach to converting a static or dynamic loop into a 3D clock assembly formed in a gel contained in a container 120. Figure 2 is a schematic diagram showing an example of the configuration of an organic gel computer (a self-learning information processing device). The organic gel computer has two orthogonal components: a microwave input and an optical output. LS is a light source, OVA is an optical vortex assembly, GT is a gel tube, NM is a nanomaterial, and PVL is a piezo-connected vortex lens. FIG. 3 shows a schematic diagram to explain how periodic or nearly periodic events in the incident data can be translated into helical nanowires in the gel-like material. 4 shows an exemplary periodic event in the input image data. From a video of a cheetah running, one of its running legs is captured and from its movement a loop is extracted for the gel computer. FIG. 5 shows a schematic diagram of a collection of spiral nanowires grown by electromagnetic fields (EM) in a container in relation to a running cheetah. FIG. 6 shows an exemplary arrangement of a container and multiple antennas, a schematic diagram of a gel-like material within the container, and an exemplary structure of nanowires grown within the container. FIG. 7 shows another exemplary configuration of nanowires grown in a container. Figure 8 shows an example of temporal images of gel material growing in a container (first column) during computation by the gel computer, corresponding SEM images taken from a slice of the R1 box region in the first column (second column), and images of in situ optical vortices projected by a self-learning information processing device. FIG. 9 shows a schematic diagram illustrating how optical vortices are formed by helical nanowires and the mathematical manipulations automated by interactive vortices. Figure 10 shows how the optical vortex hologram is analyzed for output. Panels A, B, and C show how the 2D vortex ensemble is transformed into a 3D structure. Panel D shows how the different optical rings, or optical vortices, are interconnected and their physical significance. Panel E is a zoomed-in version of Panel C, and Panel F is the output after applying a transformation function. Panel G shows the final output derived from the optical vortex. SP1 and SP2 are the two pivots around which all of the cheetah's dynamics are played out. Figure 11 shows three examples of information processing for self-learning by a gel following a complete path. Each row shows a path from input to output. The top row shows processing of a music file, the middle row shows processing of a cheetah video, and the bottom row shows processing of a swarm. FIG. 12 shows six examples of a series of big data processing by an information processing device for self-learning. FIG. 13 is a schematic diagram illustrating an information processing device for self-learning.

<<Embodiment>>
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Conceptual Overview and Aspects of the Embodiments
Before going into the details of the present embodiment, a high level overview and some aspects of the present embodiment will be briefly presented.

Factual data is either spatial, temporal, or spatiotemporal variation. Since the automaton has no knowledge of the basic elements of the three attributes (i.e., space, time, and spatiotemporal) and their variation rules in the unknown input, training of the algorithm is required.

Nature uses elemental symmetry or ordering to restrict the astronomical variation of the three attributes to a small number of components. Automata can memorize statistically dominating changes in geometry, i.e., key spatiotemporal patterns in nature.

The automaton then emulates information processing as a composition of these basic building blocks for self-learning. With a single training session, the automaton may learn rules that organize important spatiotemporal dynamics and structure unknown information. The rules memorized by the automaton sense, evaluate, and recreate all future events by integrating the basic spatial, temporal, and spatiotemporal building blocks.

This rule is an invariant hidden within a class of spatiotemporal phenomena. A young child sees a cat once, discovers the invariant, and recognizes all its variations throughout their life. The fundamental question is whether we can find the important spatiotemporal elements and a general way to discover the invariant in a single shot, reducing the need for complex and lengthy algorithms.

The demand for AI bears striking similarities to the 1919 era of geometrodynamics in astrophysics. Einstein's inseparable connection between geometric shape-shifting and time inspired and developed the idea. For a century, particle physicists and astrophysicists argued that autonomous nature describes self-learning information processing as geometric shapes hidden within the fabric of time.

Inspired by cellular automata, Feynman wrote in 1965 that the laws of physics might one day be replaced by changing geometric shapes. He tested this proposal by constructing a molecular checkerboard, in which the geometric shapes formed by similar state trajectories replicate radioactivity and diffusion, two distinct laws of physics.

However, within the last century, there has been another historic development. The idea of learning unknown data solely by mapping periodic events originated in 1941. In a century of remarkable advances, the brain's conscious experience, earthquakes, weather forecasts, deep-space astronomical events, stock markets, and viral intelligence have been theoretically mapped using structures of periodic events, or loops, known as time crystals. No historical effort has attempted to use clocks to describe geometric shapes. Inventors have found this approach possible with proteins.

Inventing an automaton that maps the structure of time in unknown data faces two problems. First, the building blocks of silicon hardware are sequential processors. Processors cannot emulate events with many different time scales, as both the scale and precision of time vary. Second, identifying overlapping loops of periodic events through unsupervised learning requires hardware that separates interconnected loops into distinct structures.

The inventors solved these two problems using an organic gel. The gel maps highly overlapping 3D arrangements of clocks in unknown data through a single, unsupervised learning process. The clocks construct the corners of geometric shapes.

Because the gel does not read or digitize the entire data, data size is not an issue. Only periodic features are sensed, creating a gel structure that vibrates like a 3D clock ensemble. As shown in Figure 1, the gel maintains three layers of temporal dynamics. First, the 3D clock ensemble. Second, the 2D planar dynamics of undefined time zones within the clock ensemble, i.e., singularities. Third, the dynamics of the lines connecting these planes. The 1D linear dynamics of the planes is independent of the spatiotemporal input. The gel learns the input as an invariant, nontrivial geometric shape. This geometric shape directly matches the invariants of the unknown geometric change. This problem is similar to the intractable clique problem, in which an algorithm must find a simple 2D geometric shape within a complex 3D geometric structure; humans can do this quickly.

By solving a wide range of database classification problems, we find that 1D, 2D, and 3D spatiotemporal inputs mapped with 4D, 5D, and 6D temporal features allow for retrieval of unencoded inputs. Extracting linear 6D identities is the first step toward reducing the role of humans in unsupervised learning.
(Configuration of information processing device 100 for self-learning)
The information processing device 100 for self-learning according to this embodiment will be described below with reference to Fig. 2. Fig. 2 is a schematic diagram of an exemplary configuration of the information processing device (device) 100 for self-learning.

The self-learning information processing device 100 may also be referred to as a self-learning device, information processing device, gel computer, or organic computer. As described below, the self-learning information processing device 100 can be considered a remotely operated, self-learning optical and magnetic vortex-based organic computer.

As shown in FIG. 2, the information processing device 100 for self-learning includes a conversion unit (first converter) 110, multiple antennas 111-1, 111-2, ..., a container 120, a laser light source 131, a pair of crossed polarizers 132, a first lens 133, a second lens 134, a first collimator 135, a first vortex lens 136, a second vortex lens 137, a second collimator 138, and a second lens 139. The information processing device 100 for self-learning may also include a microscope 141 and a monitor 142. The information processing device 100 for self-learning may also include an imaging device 151, another conversion unit (second converter) 152, and an output device 153.

In the information processing device 100 for self-learning according to this embodiment, a gel-like material is disposed in the container 120. As a specific example, the container 120 has the shape of a cylinder whose axis is parallel to the direction of gravity. The container 120 may also be referred to as a gelling tube or gel tube (GT). Examples of gel-like materials will be described later.

In the information processing device 100 for self-learning, multiple antennas 111-1, 111-2, ... are arranged around the container 120. More specifically, all antennas are arranged concentrically around the container 120 on a common horizontal base (CB). In other words, all antennas are arranged concentrically on a real or imaginary plane whose normal vector is parallel to the axis of the container 120. The multiple antennas 111-1, 111-2, ... can be collectively represented by the reference symbol 111.

In the information processing device 100 for self-learning, the conversion unit 110 receives input data and converts the input data into a set of frequency signals. The conversion unit 110 then supplies the set of frequency signals to multiple antennas 111-1, 111-2, .... The multiple antennas 111-1, 111-2, ... remotely transmit the input data to a gel material.

As a specific example, the converter 110 determines a set of frequencies for the frequency signal (input signal) using a catalog of one or more specific molecules for synthesizing helical nanowires (described below) with typical lengths, pitches, and diameters as a function of frequency. The time periods of loops in the input data vary depending on the maximum and minimum resonance times of the helical nanowire-based gel. Localized electromagnetic energy bursts in the helical nanowire assembly trigger higher-scale hierarchical self-assembly. For a particular gel material, there is an upper limit to the resonance time of each self-assembled architecture and a corresponding optical vortex with limiting values for polarization, diameter, and phase singularity around the ring.

The conversion unit 110 may be realized as a memory and a CPU (Central Processing Unit). Such a CPU can execute one or more programs to perform the above-mentioned processing.

The self-learning information processing device 100 can memorize various phenomena using periodic structures and loops, as shown below.

As shown in FIG. 2, as a specific example, the information processing device 100 for self-learning may include 12 antennas 111-1, 111-2, ..., 111-12. Each of the 12 antennas may be a Yagi antenna (Y). In this specific example, the conversion unit 110 receives input data and converts the input data into 12 frequency signals FS1, FS2, ..., FS12. The conversion unit 110 can then supply each of the 12 frequency signals FS1, FS2, ..., FS12 to each of the 12 antennas 111-1, 111-2, ..., 111-12. That is, the frequency signal FS1 is supplied to the antenna 111-1, and the frequency signal FS2 is supplied to the antenna 111-2.

As shown in FIG. 2, as a specific example, the input data may be image data. In this embodiment, the image data may be composed of multiple images or may be composed of moving images. However, this example does not limit the present embodiment. As a specific example, the input image data may include multiple pixels, each having a pixel value. That is, for example, the input image data may be divided into multiple sub-images corresponding to each horizontal scan line. Each sub-image may include multiple pixel values.

As shown in FIG. 2, as a specific example, each of the sub-images may be independently provided to the converter 110. The converter 110 may convert each of the sub-images into a frequency signal and then provide the frequency signal to a corresponding antenna. As shown in FIG. 2, for example, the converter 110 may convert the sub-image of the first scanning line into a frequency signal FS1 and then provide the frequency signal FS1 to the corresponding antenna 111-1. Similarly, the converter 110 may convert the sub-image of the second scanning line into a frequency signal FS2 and then provide the frequency signal FS2 to the corresponding antenna 111-2.

As mentioned above, image data is just one example of input data. In general, input data can be audio data, sensor data, or any other form of data.

Thus, the conversion unit 110 may be expressed as receiving each data piece of input data and then converting each data piece into a respective frequency signal in the set of frequency signals. Each frequency signal may then be provided to each of the multiple antennas 111-1, 111-2, ...

As described above, a gel-like material is placed in the container 120. The gel-like material includes one or more precursors. The one or more precursors begin to grow automatically when triggered by electromagnetic fields generated by the multiple antennas 111-1, 111-2, .... Various solvents are used to prepare a solution of the precursors that transforms into a gel when the antennas send power to the solution.

The electromagnetic fields generated by the multiple antennas 111-1, 111-2, ... grow one or more precursors into one or more clusters of helical nanowires, each of which corresponds to a periodic or nearly periodic event in the input data.

The multiple antennas 111-1, 111-2, ... are arranged to create a 3D (three-dimensional) distribution of semi-periodic electromagnetic fields, periodic electromagnetic fields, and helical electromagnetic fields within the container, where the 3D distribution of the helical electromagnetic fields includes geometric information of loops created by variables of similar magnitude in the input data. In other words, the 3D distribution of the semi-periodic electromagnetic fields, periodic electromagnetic fields, and helical electromagnetic fields includes the geometric shape of loops, where each loop represents a variable encoded in the input data.

The above aspects are explained in more detail with reference to Figure 3, which is a schematic diagram illustrating how periodic or near-periodic events in incident data can be translated into helical nanowires in a gel-like material.

As a schematic example, as shown in the first (leftmost) column of Figure 3, periodic or approximately periodic events are input as input data to the converter 110. Here, it is assumed that a hidden pendulum with periodicity T is embedded in the input data. The hidden pendulum is one example of a periodic or approximately periodic event in the input data. As described above, the converter 110 converts the input data into a set of frequency signals, and the corresponding electromagnetic fields EM are output from the multiple antennas 111.

As shown in the second (second from the left) column of FIG. 3, the gel material in the container 120 contains one or more gelator molecules (precursor) PS. The precursor PS is triggered by an electromagnetic field EM and begins to grow, as shown in the third column of FIG. 3. That is, the gelator molecules PS self-assemble in response to the electric field distribution of the electromagnetic field EM output from multiple antennas. The aggregate of precursor PS may be in the form of nanowires.

As shown in the third column of Figure 3, gravity acts on the nanowire NW1 grown in the first period. As a result, the nanowire NW1 moves downward within the container 120 along the direction of gravity, as shown in the fourth column of Figure 3. Then, during the second period, another precursor PS begins to grow within the container 120 directly above the nanowire NW1 grown in the first period, as shown in the fourth column of Figure 3.

In this manner, the precursor PS in the container 120 grows into one or more clusters of helical nanowires HNW, as shown in the fifth column of Figure 3.

After the nanowires are formed, a laser is used to retrieve the information and form an optical vortex (ring) OV, as described below. In column 6, a Poincare sphere is constructed from the optical vortex, and finally, in column 7, a pendulum is revealed. The optical vortex may also be referred to as a "clock."

From the above description, it can be understood that each cluster of helical nanowires (HNWs) corresponds to a periodic or near-periodic event in the input data.

Figure 4 shows a typical periodic event in input image data. Specifically, Figure 4 shows a time snapshot of a running cheetah. One hind leg of the cheetah is highlighted by a boundary B1 to track its movement. Here, the end point of that leg (claw) describes a roughly rectangular loop B2 that is transformed from the Cartesian coordinates in the upper row to the polar coordinates in the lower row. The non-moving position serves as the pivot. The pivot is located at the center of the circle in polar coordinates. In polar coordinates, the cheetah's hind leg is highlighted by a boundary C1, and the periodic movement of the end point of that leg is shown as loop C2.

5 shows a schematic diagram of a collection of spiral nanowires grown by an electromagnetic field EM in a container 120, where the electromagnetic field EM is converted from the image data of FIG. 4. The collection of spiral nanowires may also be referred to as a nanomaterial (NM). As shown in FIG. 5, the collection of spiral nanowires includes multiple loops of nanowires, each loop corresponding to a respective periodic portion in the input image data. More specifically, as shown in FIG. 5, the collection includes:
A loop to accommodate the cheetah's head,
A loop to accommodate the cheetah's tail,
Two loops for the cheetah's front legs, and two loops for the cheetah's back legs.

That is, the electromagnetic field converted from the image data of FIG. 4 generates a dynamic field map within the container 120. Here, the dynamic field map (FM) may be expressed as a 3D distribution of a spiral electromagnetic field, including the geometric information of loops generated by variables with similar magnitudes in the input data. In other words, the 3D distributions of the semi-periodic, periodic, and spiral electromagnetic fields include the geometric shape of loops, each representing an encoded variable in the input data. Gel fibers then fill the field map, and the gel fibers grow. The result is a nanomaterial containing nanowire loops, each corresponding to a respective portion of the input image.

Note also that the loops are intertwisted and entangled with each other, reflecting the characteristics of periodic events in the input data.

Figure 6 shows an exemplary arrangement of a container (gel tube GT) 120 and multiple antennas 111-1, 111-2, ... (leftmost part of Figure 6), a schematic diagram of the gel material in the container 120 (second from the left in Figure 6), and an exemplary structure of nanowires grown in the container 120 and captured by a microscope 141.

Figure 7 shows another exemplary configuration of nanowires grown in a container 120. More specifically, Figure 7 shows an SEM (scanning electron microscope) image of dried nanowires (dried gel (xerogel)) from the top of the container (GT) 120. A small zoomed-in portion shows a schematic of the supramolecular packing of the gelator.

As described above, according to the information processing device 100 for self-learning of this embodiment, the information contained in the input data is held in the gel material in the container 120 in the form of spiral nanowires.
(Configuration for reading out information)
Next, a configuration for reading out information held in the gel-like material in container 120 will be described. As shown in Fig. 2, information processing device 100 for self-learning includes laser light source (LS) 131, a pair of orthogonal polarizers (OP) 132, a first lens 133, a second lens 134, a first collimator 135, a first vortex lens (PVL) 136, a second vortex lens (PVL) 137, a second collimator 138, and a second lens 139. The above configuration can be regarded as a modified Fabry-Perot interferometer.

As shown in FIG. 2, a pair of vortex lenses 136 and 137 are located on opposite sides of the container 120. Vortex lens 136 is connected to a piezo resonator, and vortex lens 137 is connected to another piezo resonator.

As shown in FIG. 2, a pair of crossed polarizers 132 are positioned between the laser source 131 and the first vortex lens 136 to convert the monochromatic light into a polarized signal.

Laser light source 131 emits laser light. The laser light passes through a pair of crossed polarizers 132, a first lens 133, a second lens 134, and a first collimator 135. The laser light is then propagated as rotating light through the 3D-printed gel (grown nanowires) and reflected back and forth between two piezo oscillators attached to vortex lenses 136 and 137, respectively. The rotating light is amplified between vortex lenses 136 and 137 and projected as a collection of one or more optical vortices (OVAs) whose structure is related to the input data, as shown in FIG. 2. In other words, the pair of vortex lenses 136 and 137 project one or more optical vortices whose structure is related to the input data. In other words, a pair of vortex lenses held on two sides of the gel container reflect back and forth one or more optical vortices generated by gel nanowires and nanowire-based superstructures, whose structure preserves the geometry of periodic events in the input data. The vortex lens, connected to a piezo resonator, vibrates to resolve closely spaced vortices emanating from the gel's fiber network.

The top row of Figure 8 shows temporal images of gel-like material growing within container 120 when antenna array 111 provides an electromagnetic field corresponding to the cheetah image of Figure 4. The dotted lines in each of the top row of images in Figure 8 indicate the projected path of the vortex.

The middle row of Figure 8 shows the corresponding SEM image (scale bar: 2 μm (micrometers)) taken from a slice in the R1 box area of the top row image.

The bottom row of Figure 8 shows images (scale bar: 2 cm (centimeters)) of in situ optical vortices projected by the information processing device 100 for self-learning, corresponding to the images in the top and middle rows. Each optical vortex in the bottom row is generated from each of the R1 box regions in the image in the top row.

The top half of Figure 9 shows a schematic diagram illustrating how optical vortices are generated by helical nanowires. As shown in the top half of Figure 9, each helix (helical nanowire) forms an optical vortex. The arc differences between the dark regions on the optical ring (optical vortex) are phase gaps between interacting periodic events, forming different geometric shapes. Clusters of helices form vector vortex assemblies.

The bottom half of Figure 9 shows a schematic diagram illustrating the mathematical operations automated by interacting vortices. As shown in the bottom half of Figure 9, the mathematical operations automated by interacting vortices are shown schematically as addition and subtraction.

As described above, according to the self-learning information processing device 100 of this embodiment, the information processing for self-learning contained in the input data is held in the form of helical nanowires in the gel material within the container 120. The information held in the gel material within the container 120 is then projected in the form of one or more aggregates of one or more optical vortices.

Next, we will explain the configuration for converting optical vortices into information in the same format as the input data.

As described above, the information processing device 100 may include an imaging device 151, another conversion unit (second conversion unit) 152, and an output device 153.

The imaging device 151 captures images of optical vortices projected by the information processing device 100 for self-learning. The imaging device 151 may be a camera that captures moving images. The conversion unit 152 converts the image data captured by the imaging device into output data in a format similar to the input data using one or more conversion algorithms (conversion functions). In other words, the conversion unit 152 converts one or more optical vortex structures into output data using one or more conversion algorithms. That is, the conversion unit 152 is an optical vortex to input-like output data converter (optical vortex to input-like output data converter) that converts one or more optical vortex structures into output data using one or more conversion functions. The conversion functions are a set of successively changing geometric shapes, and the set of images is attached to the optical vortex and rotated according to the direction of the vortex polarity. By measuring the relative phase change between signal bursts emitted in two orthogonal directions by a fiber loop in the gel, a transfer function is determined, and the same transfer function is used to convert the vortex output into an input-like output.

For example, when image data is input as input data to the self-learning information processing device 100, the conversion unit 152 converts one or more optical vortex structures into output data in the form of one or more images.

The conversion unit 152 may be implemented as a CPU (Central Processing Unit) together with memory. Such a CPU can execute one or more programs to perform the above-described processing.

The output device outputs the data generated by the conversion unit 152. The output device may include a display panel or a speaker. If the input data is image data, the output device outputs the data in image format, and if the input image data is audio data, the output device outputs the data in audio format. However, the form of the output data does not limit this embodiment.

Figure 10 is a schematic diagram illustrating the process performed by the transformation unit 152. Figure 10 shows a method for retrieving information from an optical vortex. As shown in part A, an optical vortex is obtained after pumping the cheetah image data into the gel. Each spiral forms an optical vortex. Arc differences between dark regions on the light ring are assigned different phases.

Part B of Figure 10 provides a schematic presentation of an optical vortex in a phase-assigned 2D clock architecture. The clock architecture is extracted by data analysis performed by the transformation unit 152. Circles among dots represent out-of-phase light rings. Phase connectivity is indicated by nested cycles.

Part C of Figure 10 shows the preparation (generation) of a Poincaré sphere from a 2D clock representing electromagnetic vector directions in 3D space. The Poincaré sphere may be generated by the transformation unit 152.

As shown in part D of Figure 10, a fractal map is constructed based on the optical vortex to differentiate any similar dynamics of other animals. The fractal map may be generated by the transform unit 152. As shown in part D of Figure 10, as a specific example, the fractal map starts with large cycles (φ) and gradually progresses to smaller cycles. All body parts are represented by a set of circles below the horizontal line. These circles can represent the body parts in the formed vortex of Figure 5.

In part D of Figure 10, the Greek letters indicate the topology of a particular circle, and the numbers indicate the number of black dots on the ring. The end point at the end of the fractal chart, indicated by the arrow, does not exist in the case of the lioness.

Next, as shown in part E of Figure 10, a 3D sphere is formed by the transform unit 152 based on the 2D clock architecture. When multiple cycles merge there together, two singular points SP1 and SP2 are generated.

Next, as shown in part F of Figure 10, static points are generated by running a clock around the singular point. The static points are generated by the conversion unit 152. Circle C1 and small circles of the same tone within and near circle C1 represent the hind legs. Circle C2 and small circles of the same tone within and near circle C2 represent the front legs. Circle C3 and small circles of the same tone within and near circle C3 represent the torso. Circle C4 and small circles of the same tone within and near circle C4 represent the head. Circle C5 and small circles of the same tone within and near circle C5 represent the tail.

From singular points SP1 and SP2, static points are connected by transformer 152 to figure out body parts at a given time. Connections between similar points form dynamic loops of body organs. All circles are finally hosted by a larger circle with corresponding shading, which confines the limits of their motion.

Finally, as shown in part G of Figure 10, by drawing an ellipsoid in a teardrop shape and centering it at singular points SP1 and SP2, the final structure of the cheetah is reconstructed and retrieved by the transformation unit 152.

Figure 11 shows another example to illustrate the process performed by the transform unit 152. The three rows in Figure 11 show the reconstruction of input data from the output vortex collection. The top row A shows the case where the input data is music. The middle row B shows the case where the input data is image data of a cheetah. The bottom row C shows the case where the input data is image data of starling dynamics.

Figure 11 has five columns. The first column shows the input frame. The first column in column A shows the sound intensity versus time profile. The second column in columns A, B, and C shows the output optical vortex generated in the interferometer (pair of vortex lenses 136, 137) of the information processing device 100 for self-learning.

The third column shows an approximate circular or clock representation for the vortex output of the second column. The circular representation, i.e., the clock representation, may be derived by the converter 152. The first type spots and second type spots in the third column represent singular points and system points of the clock structure, respectively.

In the fourth column, small clusters of circles or clocks from the third column are separated and stereographically projected onto lines (column A) or areas (columns B and C). In column A, the clock raster projects Gaussian peaks on both sides of the X axis. The scale and Hz of the fixed frequencies are "Ni-Re-Ga (C4: 277 Hz, E4: 329 Hz, F4: 369 Hz)" and "Pa-Re (A4: 440 Hz, E4: 329 Hz)." In columns B and C, the first type of spot forms the pivot around which the second type of spot moves. The transformer 152 traces shapes ranging from teardrops to ellipses in the fourth column. The second type of spot and the first type of spot are connected to recreate the actual input data in the fifth column.

Figure 12 shows a series of big data processing steps performed by the information processing device 100 for self-learning. Six columns show the step-by-step processing of big data. Here, the first column shows the moments captured in the big data. The second column shows a geometric analysis of the data that attributes vortex formation. The third column shows an analysis of the phase of the optical vortex.

Column 4 shows the fractal chart, i.e., the phase correlation of the periodic loops of the big data set. Column 5 shows the Poincaré sphere derived from the optical vortex. Column 6 shows the self-learning retrieval by information processing from the optical vortex. All static points in column 6 are represented by spots of the second type, and singular points are represented by spots of the first type.

In Figure 12, row A shows how a running cheetah is retrieved. Row B shows how to pour water into a pot. Row C shows the fourth subtype of diabetes data from the input file. Row D shows how the human ACE2 receptor interacts with SARS-CoV2 spike proteins corresponding to different mammals, and how the evolution of a stronger interaction predicts a more infectious SARS-CoV virus. Row E shows the dynamics of starlings, divided into four groups, interpreting information from insects. Row F shows how the information processing device for self-learning 100 derives classes of Indian classical music from the audio track of a movie song.

(Information processing methods for self-study)
The information processing method for self-learning according to this embodiment will be described below. The information processing method for self-learning according to this embodiment may include the following steps.

(Step S101)
In step S101, a gel material placed in a container 120 is prepared.

(Step S102)
In step S102, an electromagnetic field corresponding to the input data is generated by the conversion unit 110. The electromagnetic field is supplied to a plurality of antennas 111-1, 111-2, . . . arranged around the container 120.

In other words, the information processing method for self-learning is:
preparing a neural network-like helical nanowire-based fiber that builds a fractal-like superstructure within a container, where the output product of the one-step synthesis is used as starting material for the next step;
generating loops of intertwined electromagnetic fields, which generate loops of all variables in the input data and trigger self-organization on the order of a million, from atomic to visible scales, synthesizing an integrated self-organizing architecture of the clock;
It has.

This information processing method makes it possible to memorize various phenomena using periodic structures and loops.

Note that the information processing method for self-learning may include other steps performed by the information processing device 100 for self-learning described in the specification.

The information processing method for self-learning may also be referred to as a self-learning method, i.e., an information processing method. As described below, the information processing method for self-learning may also be referred to as a remotely operated, self-learning, optical vortex/magnetic vortex-based organic computing method.
(Further description of the information processing device 100 for self-learning)
The following further describes the self-learning information processing device 100. Fig. 13 shows how the self-learning information processing device 100 rewrites big data as a composition of optical knots and vortices.

In part A of Figure 13, the rapidly changing dynamics of a running cheetah are plotted in five shades of gray. When all bits are acquired, they become big data. The statistical database of the four periodic motion loops corresponding to the hind legs may have astronomical correlations. The overlapping loop regions may be represented as confusion, indicated by the arrows. As shown in part B of Figure 13, the few dominant loops in the big data of part A form distinct geometric spirals in part B of Figure 13. Each spiral forms an optical vortex. In part C of Figure 13, a cluster of spirals forms a vector vortex assembly. The cluster operates as a unit and interacts with the construct scalar interference loop of the dark 3D line (dark knot). Both the vector vortex and the scalar vortex converge on the Poincaré sphere. In part D of Figure 13, the arc differences between the dark regions on the optical ring are the phase gaps between the interacting periodic events shown in part A of Figure 13. The three confusion loops in part A form three singular points in the pentagon in part D. The "pentagon + square" in part D is the architecture of the confusion.

As mentioned in the "Conceptual Overview and Aspects of the Embodiments" section and above, the gel does not read or digitize the entire data, so data size is not an issue. Only periodic features are sensed, creating a gel structure that vibrates like a 3D clock ensemble. As shown in Figure 1, the gel preserves three layers of temporal dynamics. First, a 3D ensemble of clocks. Second, 2D planar dynamics of undefined time zones, or singularities, within the clock ensemble. Third, dynamics of the lines connecting these planes. The 1D linear dynamics of the planes is independent of the spatiotemporal input. The gel learns the spatiotemporal input as a key geometric invariant. Its geometry directly matches the invariant of the unknown geometric change. This problem is similar to the intractable clique problem. In the clique problem, an algorithm must find a simple 2D geometric shape within a complex 3D geometric structure, something humans can do quickly.

As described above, by solving a wide range of database classification problems, we have found that 1D, 2D, and 3D spatiotemporal inputs mapped with 4D, 5D, and 6D temporal features enable the unencoded inputs to be retrieved. Extracting linear 6D identities is the first step toward reducing the role of humans in unsupervised learning.

As shown in Figure 3, the GEL setup, an example of a self-learning information processing device 100, acts as a 3D arrangement of clocks to instantly identify, track, and map periodic events in big data. A cavity resonator (pair of vortex lenses 136 and 137) is used to synthesize one solid, transparent helical nanowire at each periodic event. The resonating nanowire edits the input rotating light, i.e., an optical vortex. Depending on the length, pitch, and diameter of the helix, one to six phase singularities form on the perimeter of the optical ring. As shown in the upper half of Figure 9, the singularities appear as dark spots. Arc gaps between the dark spots clarify the geometrical links between sub-events.

The gel structure within vessel 120 is translucent and allows light and signal frequencies to pass through, causing the field of helical paths to interact with the gel molecules. As a result, the input monochromatic laser light used to read the helices generates a collection of light rings whose angular momentum changes in proportion to the geometric parameters of each distinct helix synthesized within the chemical pot (vessel 120), creating a hologram of light rings that are projected.

As events grow within and above the input data (unknown data), the spirals combine within the chemical beaker (vessel 120), and the emitted light rings also combine as a unit. The diameter of the reflected light rings expands or contracts according to the periodicity of the events. As scaling up, the precisely oriented spirals resonate, and the rapidly synthesized nanowire ensemble emulates all possible phase relationships between all the nearly periodic events in the input data. The complex links between events are encoded in the emitted hologram (optical vortex) as dark spots, light ring diameters, rotating light directions, and relative 3D orientations of the 2D planar light rings. The holographic 3D arrangement of the light rings live-feeds precise changes in the big data. As new intelligent features emerge in the big data, new light rings appear. As the relationships between periodic events change, the rings instantly shift on the hologram.

As previously explained, the inventors optimized the setup shown in Figure 2 to ensure that the relationships between periodic events hidden in the input data (unknown data) are accurately encoded in the synthesized nanowires and projected holograms. The setup has two orthogonal sections. The vertical section is a 3D printer that live-feeds preprocessed big data into the synthesis beaker (GT) (container 120) using an antenna array (antennas 111-1, 111-2, ...) arranged around the entire GT. Preprocessing by the converter 110 converts the input data into a 2D or 3D matrix of microwave frequencies that identify organic precursors. The axes of the input 2D or 3D matrix are correlation variables. These are converted to (r, θ) or (r, θ, φ) in the hologram output, making it straightforward to revert to the input format.

For wireless input, the input value is blindly altered to electromagnetic frequencies (EM). The antenna array feed builds a 2D disk of EM field pattern within the GT. Gelator molecules in reservoir 120 resonate and fill the favored field lines. The printed disk descends due to gravity. By this time, the antenna array has updated data—new disk shapes, new molecules—and pours them into empty field cages. Synthesis continues disk by disk. Iso-frequency paths across the disk build transparent spiral nanowires.

The horizontal portion of the setup in Figure 2 is a modified Fabry-Perot interferometer that sends rotating light through a 3D-printed gel and reflects it back and forth between two piezo oscillators attached to vortex lenses 136 and 137, respectively. Vortex lenses 136 and 137 are used to induce angular momentum in the input 633 nm He-Ne monochromatic laser light, creating an optical ring, or optical vortex.

The repeated reflections perform two operations: First, they amplify the weak 2D light rings by an order of magnitude of 10 ⁵ and superimpose them onto a phase sphere with light rings all around it, understood as a 3D hologram.

The gel's morphology, frequency range, and antenna-cavity geometry are key to the one-to-one correspondence between periodic events ψ_i (psi_i) in the input data, the helices generated in the gel, and the optical ring ensembles in the holographic output. The gelator, (S)-phenyl-tetradecanoylamino-acetic acid methyl ester, spontaneously forms a gel of randomly shaped helical nanowires in a heptane solution (sol). The antenna array precisely edits the length, pitch, and diameter of the helical nanowires, corresponding to each pixel of the big data. Wirelessly transmitted EM frequencies generate helices of specific geometric shapes. At higher frequencies (MHz), atomic-scale helices are constructed, and these helices self-organize into macrohelices that resonate at much lower frequencies (kHz, Hz). Once written, the solid gel retains the complex 3D orientation of the nanohelices for many years. The gel melts at 70°C, ready for new input. There is no need to replace the gel in the GT; the gel can be melted and reused.

The confusion at the top automatically connects to the facts below it in the confusion structure. This principle may be called the "Principle of Doubt (D2)." Principle D2 does not resolve confusions, but maps both facts and confusions into a 3D hierarchical net. D2 appears to be a deep learning net, providing all possible connections as answers to any question asked by connecting the confusions, i.e., facts, with lines on D2. If D2 is formed from 40 confusions, it has the potential to address ^2-40 questions.

In some aspects, the self-learning information processing device 100 may be considered a D2 computer. To operate the self-learning information processing device 100, physical attributes for facts and confusion states are required. The periodic events of big data (Equation (1)) are equivalent to the rotation of an electric vector around the circumference of a light ring.

Here, l (Equation (3)) is the angular momentum, and Φ (Equation (4)) is the rotation phase.

The vortices interfere, generating dark spots around the ring. In the optical ring, i.e., in the n dark spots on the optical vortex, we obtain the confusion state (Equation (5)).

For n ₂ =1 (equation (6)), we obtain the fact, and for n ₂ ≧2 (equation (7)), we obtain the confusion state, which is true for scalar vortices.

In vector vortices, _n2 (Eq. (8)) is the number of classes of torsion of the ring, since in vector vortices the number of dark spots can be sufficiently large. Each facet of the confusion architecture contains multiple _Ψi (Eq. (9)).

The vortices may be conditionally coupled to form a confused ring, i.e., (Equation (10)) is obtained.

Multiple clusters of ψ _i (Eq. (11)) form a layer of 1. The confused architecture grows within and above, so if there are n ₁ (Eq. (12)) layers, the wave function representing the confused architecture D2 is (Eq. (13)).

To solve the problem "n ₁ ≧3" (Eq. (14)), the product of decision Ψ _ac (Eq. (15)) is viewed as a collection of optical vortices. The equivalent nested phase spheres that reveal the 3D relationships between periodic events in big data are the intelligence of the system.

To briefly explain D2 behavior, we mapped all the periodic events Ψ _ac (Equation (16)) of a running cheetah and constructed D2. The cheetah learning gel analyzed a running lioness and immediately suggested key dynamic differences.

As shown in Figure 2, while the video input of the cheetah was pumped into the gelator solution, the cheetah's rectangular frame was sliced into 12 rows and 22 columns using 12 Yagi antennas (111-1, ..., 111-12) arranged all around the periphery of the GT (container 120), and 12 pixels per column were supplied at a time.

By solving Maxwell's equations, we determined the distribution of the electromagnetic field when fed an image of a cheetah. A column of ones forms a field disk, a frame of ones forms a unit cylinder of the field, and the image increases the length of the field cylinder. If a pixel in the image returns to its starting point, we get one period of the spiral in GT. As the cheetah runs, a new pitch is added to the spiral, and the gel fills the field container. The image of a running cheetah contains many pixels moving in a loop.

However, because the 3D interference of the EM field within the cavity combines non-planar loops, the astronomical number of self-similar loops from a particular organ shrinks to one. As shown in Figure 4, while running, the cheetah's two hind legs and tail trace three loops with a resting point at the hips. Similarly, the two front legs and head trace three loops with a resting point at the shoulders. Two static points connect the centers of the polar plots. SEM images show that the GT cavity reliably transforms the nested loops into nested spirals of the desired geometry.

Transformer 152 allows intuitive conversion of the gel-projected vortex hologram into a cognitive format. However, a simple algorithm, i.e., a set of operations, produces an output similar to the input. Transformer 152 counts the number of vortices in the hologram, the number of dark spots on each ring, and angular momentum sorting reveals the relative phase of all rings.

The diameter of the ring is r_n, the in-plane phase is θ_n, and the angle between the non-planar clocks is φ_n. The relative phase underpins the location of the clock's starting system point. Thus, with the transformation 152, the 3D clock ensemble is obtained as a 4D dynamics. The inventors have noticed that the delay of the phase singularity domain, i.e., the delay of the dark spot, which limits the nested clocks, terminates the route.

Next, the transformer 152 finds periodically oscillating lines consisting of undefined time zones. These lines are the 2D global clocks that drive the nested 3D clocks within and above the network. The phase singularities act as axes, driving the clocks in the 2D plane. We call it 5D because we need a new orthogonal axis to the 4D time scale. This plane remains constant under 3D spherical variations in each clock, and the collection of clocks represents a similar hologram. The 5D planes are connected by periodically oscillating threads, but their linear dynamics are insensitive to plane shifts, so we add a new orthogonal axis to the 5D and assign 1D lines to the 6D dynamics.

The 1D, 2D, and 3D time-structure-derived collection of circles appears as a superposition of all frames of the running cheetah. Transformer 152 scales many circles as the output of a small number of geometric shape transitions relative to the axis of the pendulum. As the clock moves, the 3D clock collection in the hologram generates an infinite number of future running steps for the cheetah. The 1D and 2D invariants remain constant even when the gel is fed with 30 images of the cheetah.

Invariant detection is universal. Figure 12 lists the widely varied data formats read and analyzed by the gel. The inventors fed the gel diabetes gene expression data already elucidated by deep learning. The hologram revealed hidden genetic expressions, just as deep learning predicted. Thus, the gel retrieves periods in static data. The inventors fed the gel the shapes of coronavirus spikes and cell receptor ACE2 junctions, discovering intelligence used by animals to evolve their structures.

This hologram shows that the coronavirus has been perfecting its connections for 18 years, making it more infectious. Thus, the gel filters out randomly mixed periods. Bird families are destroyed and new groups are created while flying through the sky. The gel finds enough group regularities in the image that if an eagle attacks the group, the gel predicts a panicked response. Thus, even if the input pattern randomly disappears, the gel's periodic mapping survives.

A pure audio file of frequency packets arranged in a time loop is encoded into the gel by single-shot learning. A complex song embedded in the frequency pattern is fed to the gel. The gel extracts memorized patterns in the complex input. Thus, it solves the clique problem. Finally, the inventors took on the open challenge of classifying random videos of liquid dripping from one pot to another. The gel classifies variations in direction, position, and velocity after single-shot learning. Thus, the gel memorizes key 1D and 2D invariants from any input format and finds them by resonance in any input 1D, 2D, or 3D pattern.

Because the learned gel resonates with the 1D and 2D invariants, the response reaches the interrogator on its own, without the need to explore or construct a circuit. By locking five space-time transformations on the 1D and 2D invariants and implementing a 3D clock, we can reconstruct output 1D, 2D, 3D spatial, temporal, and spatiotemporal patterns that resemble the input with 86% accuracy.

When light rings are rewritten as nested phase spheres, a vast number of astronomical paths are preserved that are never encoded in the gel. Each path acts as an algorithm. The algorithm detects periodic events as physical movements of pixels; therefore, it cannot discover conditional phase relationships between events. The gel discovers conditional phase relationships because it discovers 1D and 2D invariants in the time domain. The gel holds a holographic key; as light rotates, geometric shapes are actuated.

The gel nanowires generate holograms based on the symmetry of the nanowire array. The hologram is insensitive to the number of nanowires or to imperfections in the nanowire-made lattice. Therefore, if five guest clocks embedded in a host clock exhibit dynamics, the pentagon can deform in enormous ways within the clock's input 3D geometry. (Note that we use hierarchical terminology here, where "guest" clocks can be embedded in a "host" clock. We also use the terms "daughter" and "mother" clocks to represent the clock hierarchy.) The gel still generates a typical pentagonal hologram. Furthermore, the clocks grow within and above it, meaning that there are geometric shapes inside the corners of the geometric shapes.

Thus, every cheetah runs differently, and precisely, the 1D and 2D invariants of different running styles accurately underpin. Finally, when closely approximated geometric compositions are applied to the learning gel, the output light ring quantitatively reveals similarities and differences in the preserved dynamics.

In real-world data, the coupling conditions of nearly periodic events change over time, loop shapes fluctuate, and loops overlap and regroup. The language of GEL is learning geometry using basic geometry. Notably, moving spatially to higher dimensions starts in 1D, and moving temporally ends in 6D, yet with 1D dynamics.

Thus, loops reshape astronomically, but 1D invariants preserve the conformal vortex map of the gel. Therefore, the strength of organic gel processing is finding and precisely orienting dense clocks. As a result, periodicity mapping shifts focus from the half-century-old demands of ever-increasing memory capacity and data processing speed to the demands of time bandwidth, where the gel could intricately retrieve a 3D clock map, i.e., invariants, from unknown data. The gel invariant extraction time is independent of data size. Fast and slow gels can be combined as needed.

The time domain into which the gel resolves nested events determines its efficiency, precision, and predictability. When a very slow clock hosts a faster periodic clock in a 3D clock ensemble, non-repeating events can unfold over centuries. The inventors mapped all brain rhythms onto a unique 3D clock architecture as part of a self-operating mathematical universe (SOMU) that learns to operate quickly, without algorithms.
(Notes about the D2 protocol)
As partially described above, the concept of "D2" may be introduced behind the realization of the information processing device 100 for self-learning. The D2 protocol (D2 algorithm) can be expressed as follows:

Instead of finding facts, the D2 protocol searches for confusions, i.e., unclear concepts. Then, within the confusions, the D2 protocol attempts to find more confusions. The search continues until it encounters a fact. The 3D structure obtained by the following steps may be called the confusion architecture. Either all events, i.e., facts, are cyclic, or events are periodic. A confusion is a superposition of many periodic events. The confusion architecture is represented using a phase sphere, on whose surface there is a circle representing a clock. The system point moves on this circle. The 3D clock representation of the confusion architecture reveals all possible phase relationships between all periodic events in the unknown database, without reading them. This is the D2 protocol.

<<Various Aspects of the Embodiments>>
The embodiments have various aspects. The aspects and explanations described in the embodiments can be expressed as follows: The information processing device 100 for self-learning can be regarded as an analog real-time unknown data (input data) analyzer.

<Aspect 1-1>
An analog instant unknown data analyzer consisting of a 3D printer (transducer 110, an array of antennas 111, and a container 120) converts periodic or near-periodic events in unknown data (input data) into a collection of spiral nanowires, irradiates polarized light (by a laser light source 131), and converts (by a converter 152) a projected hologram generated by the instant nanowire collection with respect to the input data.

<Aspect 1-2>
In the analog real-time unknown data analyzer according to aspect 1-1, the pre-processing algorithm (executed by the conversion unit 110) incorporates static or dynamic data (input data) expressed in terms of variables into a set of frequencies using a catalog of specific molecules that synthesize helical nanowires with typical lengths, pitches, and diameters as a function of the number of turns.
<Aspect 1-3>
In the analog real-time unknown data analyzer described in aspect 1-1 or 1-2, the 3D antenna array (antennas 111-1, 11-2, ...) and the gel synthesis pot (container 120) are configured together to generate a 3D distribution of a spiral field (spiral EM field), the field distribution including geometric parameters of a loop generated by variables having magnitudes approximating the unknown data (input data).

<Aspect 1-4>
An analog real-time unknown data analyzer according to any one of Aspects 1-1 to 1-3, wherein a composition, i.e., a singular molecular gel precursor, is held in a gel synthesis pot (container 120) and begins to grow from a molecular scale to a scale similar to the dimension of the filled distribution, and has the property of filling an iso-frequency helical field region.
<Aspect 1-5>
An analog real-time unknown data analyzer according to any one of aspects 1-1 to 1-4, wherein the gel structure is semi-transparent and passes light and a signal frequency that causes a field of helical paths to interact with the gel molecules, such that the input monochromatic light that reads the helices changes angular momentum in proportion to the geometric parameters of each individual helix synthesized in the chemical pot, generating a collection of projected light rings as a hologram.
<Aspect 1-6>
An analog real-time unknown data analyzer according to any one of aspects 1-1 to 1-5, wherein one or a pair of vortex lenses (136 and 137) connected to a piezo resonator having a hole are held on two sides of the chemical synthesis pot (the container 120) along the light beam, and the pair of resonantly vibrating piezo-connected vortex lenses amplifies and scans the frequency spectrum of the spiral nanowire assembly by vibration.
<Aspect 1-7>
In the analog real-time unknown data analyzer according to any one of Aspects 1-1 to 1-6, for post-processing of the output vortices (performed by the conversion unit 152), a set of conversion functions is applied to the vortex ring ensemble to convert the output vortices into a static or dynamic pattern, which appears as a variable plot that is sent as input data to the helical nanowire ensemble solution (the gelatinous material in the container 120).
<Aspect 2-1>
An analog real-time unknown data analyzer according to any one of aspects 1-1 to 1-7, which uses an algorithm in a simulator (a simulator including an information processing device 100 for self-learning) to detect loops of similar values in input 1D, 2D, and 3D variables, convert the loops directly into a hologram of an optical vortex as output, and repeatedly convert the output into a plot resembling the input variables.
<Aspect 2-2>
In the analog real-time unknown data analyzer of aspect 2-1, the relative spatial and temporal relationships between interconnected loops of like values are determined within a vessel 120 to form a 3D plot of phase spheres, each phase sphere formed for one loop, with circles on the sphere representing the loops, with system points placed around them and motion assigned to the system points to represent the dynamics of the loops of like values in the unknown data.
<Aspect 2-3>
In an analog real-time unknown data analyzer according to aspect 2-1 or 2-2, a collection of 3D phase spheres generated from a loop of similar values in the unknown data (input data) functions as a guest-host system, where one phase sphere is inserted into another phase sphere, and the other phase sphere selects a location for an undefined phase region having a boundary similar to that of the inserted phase sphere.
<Aspect 2-4>
In the analog real-time unknown data analyzer according to any one of Aspects 2-1 to 2-3, the phases, angular momentum, and rotation directions of all rotating circles that function like a clock within the 3D phase sphere assembly function as memory elements that the data analyzer (information processing device 100 for self-learning) learns from, and when new unknown data is sent as input to the simulator (information processing device 100 for self-learning), similarities and differences in various symmetries are determined.
<Aspect 3-1>
The pre-processing and post-processing algorithms of the Analog Real-Time Unknown Data Analyzer run extensive simulations in which no near-periodic events are found, and through mathematical transformations, the database is iterated to find periodicity in said events.
<Aspect 3-2>
In the pre-processing and post-processing algorithms of the analog real-time unknown data analyzer described in aspect 3-1, for pre-processing of semi-periodic events, if a variable does not complete a time loop, jumps from one point to another, crosses its own loop, a mean path is assumed, and if a system point moves periodically along a line in the unknown data, it is considered a loop and a new phase sphere is formed.

<Aspect 3-3>
A pre-processing algorithm and a post-processing algorithm for an analog real-time unknown data analyzer according to aspect 3-1 or 3-2, wherein loops in the unknown data where similar values overlap are superimposed on a singular sphere, and loops having a common region generate a common phase sphere where they overlap within the common region.
<Aspect 3-4>
A pre-processing algorithm and a post-processing algorithm for an analog real-time unknown data analyzer according to any one of aspects 3-1 to 3-3, wherein 15 selected transformation functions are applied to input data transformed into a 3D phase sphere ensemble to find the best composition of correlation functions that most accurately transforms the input to the 3D phase sphere ensemble, and the methodologies identified for the transformation during pre-processing are used when the output vortex ensemble or sphere ensemble is transformed into an output like the input via post-processing.
<Aspect 4-1>
An analog real-time unknown data analyzer in which a chemical synthesis pot surrounded by an electromagnetic antenna is configured to construct a 3D architecture of looped helical nanowires via supramolecular assembly.
<Aspect 4-2>
An analog real-time unknown data analyzer as described in aspect 4-1, in which unknown data is divided into pixels, normalized to a set value, and an electromagnetic frequency catalog is applied, and each pixel is associated with the typical electromagnetic frequency and reflection/transmission/absorption of a chemical synthesis pot (container 120), and the nanowire synthesis frequencies from the precursors do not affect each other and are kept in isolated bands.
<Aspect 4-3>
An analog real-time unknown data analyzer according to aspect 4-1 or 4-2, wherein the boundary of the chemical pot (container 120) acts as a dielectric resonator, and the cavity acts as a cavity resonator, i.e., the material, and in each substantially periodic loop, a helical distribution field is formed, 2D angular changes form 3D angular orientations, and tangled loops form tangled helices.
<Aspect 4-4>
An analog real-time unknown data analyzer according to any one of aspects 4-1 or 4-3, wherein the array antennas (antennas 111-1, 111-2, ...) are arranged in one or more ring-like arrangements in the shape of a ring, i.e., a column of rings, around the entire circumference of the chemical synthesis pot (container 120), thereby forming a columnar region of field distribution, thereby ensuring multiple replicas of the helical nanowire assembly.
<Aspect 5-1>
In an analog real-time unknown data analyzer, piezo connected vortex lenses (136 and 137) form a hologram made of modified vortices.
<Aspect 5-2>
An analog real-time unknown data analyzer according to aspect 5-1, wherein the vibration of the piezoelectric resonator is optimized for the vortex lens, the piezoelectric-coupled vortex lens resonates at a desired frequency sufficient to change the wavelength used for helical nanowire synthesis, and the vibration is tuned to the applied laser read frequency to progress a frequency scan.
<Aspect 5-3>
In the analog real-time unknown data analyzer according to aspect 5-1 or 5-2, the vortex lenses (136 and 137) rotate the light and add angular momentum to the reading light, and repeated reflections between a pair of piezo-coupled vortex lenses increase the scanning frequency bandwidth and frequency resolution of the scan for reading information of the helical nanowire ensemble.
<Aspect 5-4>
An analog real-time unknown data analyzer according to any one of aspects 5-1 to 5-3, wherein the piezo resonator connected to the vortex lenses (136 and 137) has a hole that allows light to pass through and reach the outside as an output that is ultimately read by a semiconductor camera, and the relative angle between the two lenses (136 and 137) is adjusted relative to each other and the chemical synthesis pot (the container 120), and an integrated 3D vortex ensemble is emitted outside from the spiral nanowire ensemble.
<Aspect 6-1>
In an analog real-time unknown data analyzer, a molecular precursor is selected that forms helical nanowires of a wide range of geometric parameters as a function of applied electromagnetic frequency.
<Aspect 6-2>
An analog real-time unknown data analyzer according to aspect 6-1, wherein the reaction kinetics rate of chemical synthesis from precursors to helical nanowires matches the rate at which data is supplied as electromagnetic signals per pixel to the chemical synthesis cavity by the antenna, and any mismatch results in discontinuities in the structure.
<Aspect 6-3>
An analog real-time unknown data analyzer according to aspect 6-1 or 6-2, wherein the suitable molecular precursor, i.e., any synthetic part, forms transparent helical nanowires with appropriate length, pitch, and diameter as a function of applied electromagnetic frequency, and generates multiple replicas of its derivative to build transparent superstructures that do not rapidly settle and float against gravity.
<Aspect 6-4>
An analog real-time unknown data analyzer according to any one of aspects 6-1 to 6-3, wherein one or more molecular precursors, i.e., any composition of spiral nanowire synthesis units, are used to cover a wide range of loops found in any unknown data (input data) and accommodate a wide variation in the rate at which data is updated when unknown data is supplied to the chemical synthesis pot (vessel 120).
[Additional Note 1]
The present invention is not limited to the above exemplary embodiments, and can be modified in various ways by those skilled in the art within the scope of the claims. For example, the present invention encompasses exemplary embodiments obtained by appropriate combinations of the technical means disclosed in the above exemplary embodiments within its technical scope.
[Additional Note 2]
The entire or part of the exemplary embodiments disclosed above will be described below. However, it should be noted that the present invention is not limited to the aspects of the following examples.
(Aspect A1)
An information processing device for self-learning,
A container and
a gel material disposed within the container;
a plurality of antennas disposed around the container, the antennas generating electromagnetic fields corresponding to input data;
An information processing device comprising:
(Aspect A2)
The information processing device for self-learning according to aspect A1,
The information processing device, wherein the plurality of antennas are concentrically arranged around the container on a common horizontal base.
(Aspect A3)
The information processing device for self-learning according to aspect 1 or aspect 2,
a first converter for converting input data into a set of frequency signals;
The set of frequency signals is supplied to the plurality of antennas.
(Aspect A4)
The information processing device for self-learning according to any one of Aspects A1 to A3,
The gel material includes one or more precursors that begin to grow when triggered by an electromagnetic field formed by the plurality of antennas.
(Aspect A5)
The information processing device for self-learning according to aspect A4,
the electromagnetic field formed by the plurality of antennas grows the one or more precursors into one or more collections of helical nanowires;
An information processing device, wherein each of the one or more collections of helical nanowires corresponds to a periodic or approximately periodic event in the input data.
(Aspect A6)
An information processing device for self-learning according to any one of aspects A1 to A5,
the plurality of antennas are arranged such that the plurality of antennas generate a 3D distribution of a spiral electromagnetic field;
10. An information processing apparatus, comprising: a geometric self-learning by loop information processing, wherein the 3D distribution of the spiral electromagnetic field is formed by variables having similar magnitudes in the input data.
(Aspect A7)
The information processing device for self-learning according to any one of Aspects A1 to A6,
An information processing device, wherein the frequencies of the set of frequency signals are determined using a catalog of one or more specific molecules that synthesize helical nanowires having typical lengths, pitches, and diameters as a function of frequency.
(Aspect A8)
An information processing device for self-learning according to any one of aspects A1 to A7,
a laser light source;
a pair of vortex lenses, one on each side of the vessel;
and
An information processing device, wherein each of the vortex lenses is connected to one piezoelectric resonator.
(Aspect A9)
The information processing device for self-learning according to aspect A8,
The information processing device further comprising a pair of crossed polarizers positioned between the laser light source and one of the vortex lenses.
(Aspect A10)
The information processing device for self-learning according to any one of Aspects A1 to A9,
The pair of vortex lenses projects one or more optical vortices having a structure related to prior input data.
(Aspect A11)
The information processing device for self-study according to aspect A10,
The information processing device further comprises a second transformer that transforms the one or more optical vortex structures into output data using one or more transformation algorithms.
(Aspect A12)
An information processing method for self-learning, comprising:
Providing a gel material disposed within a container;
generating an electromagnetic field corresponding to input data and applying the electromagnetic field to a plurality of antennas disposed around the container;
An information processing method comprising:

100 Information processing device for self-learning 110 Conversion unit (first converter)
111 Antenna 120 Container 131 Laser light source 132 Crossed polarizer 136 First vortex lens 137 Second vortex lens 152 Converter (second converter)

Claims (11)

A gel processing apparatus for processing a gel-like material , comprising:
A container and
a gel material disposed within the container;
a plurality of antennas arranged around the container to generate electromagnetic fields corresponding to input data;
a pair of vortex lenses connected to a laser light source and a piezoelectric resonator;
Equipped with
the gel material is self-assembled by the electromagnetic fields from the plurality of antennas;
The laser light emitted from the laser light source passes through one of the pair of vortex lenses and is irradiated onto the self-assembled gel-like material, and an optical vortex that reflects the structure of the gel-like material is formed from the other of the pair of vortex lenses.
Gel processing equipment. 2. The gel processing device according to claim 1,
A gel processing device, wherein the plurality of antennas are concentrically arranged around the container on a common horizontal base. 3. The gel processing device according to claim 1 or 2,
a first converter for converting the input data into a set of frequency signals;
The set of frequency signals is provided to the plurality of antennas, which remotely transmit the input data to the gel material . The gel processing device according to any one of claims 1 to 3,
the gel material includes one or more precursors whose growth is triggered by an electromagnetic field formed by the plurality of antennas;
A gel processing device, wherein various solvents are used to prepare the precursor solution, and the precursor solution is transformed into a gel by the plurality of antennas transmitting the input data to the solution. 5. The gel processing device according to claim 4,
the electromagnetic field formed by the plurality of antennas grows the one or more precursors into one or more collections of helical nanowires;
The gel processing device, wherein each of the one or more collections of helical nanowires corresponds to a periodic or near-periodic event in the input data. 6. The gel processing device according to claim 1,
the plurality of antennas are arranged to radiate a 3D distribution of a spiral electromagnetic field within the vessel;
The gel processing device , wherein the 3D distribution of a semi-periodic, periodic, or helical electromagnetic field comprises a geometry of loops, each loop representing a variable encoded in the input data. 7. The gel processing device according to claim 1,
the frequency of the input data is determined using a catalog of one or more specific molecules that synthesize helical nanowires as a function of frequency;
the time period of the loop in the input data varies according to the maximum and minimum times of higher order resonances of the helical nanowire-based gel material;
a localized electromagnetic field at the collection of helical nanowires triggers higher order hierarchical self-assembly;
A gel processing device in which for a particular gel-like material, there exists a corresponding optical vortex with upper time limits for resonance of each self-assembled architecture and limiting values for polarization, diameter, and phase singularity on the ring circumference. 8. The gel processing device according to claim 1,
The gel processing device further comprising a pair of crossed polarizers disposed between the laser light source and one of the plurality of vortex lenses to convert the monochromatic light into a polarized light signal. 9. The gel processing device according to claim 1,
the pair of vortex lenses are held on two sides of the container and reflect back and forth one or more optical vortices generated by gel nanowires and nanowire-based superstructures, the structure of which preserves the geometry of periodic events in the input data;
A vortex lens connected to the piezo resonator vibrates to separate adjacent vortices emanating from a fiber network of the gel material. 10. The gel processing device according to claim 9,
a data converter that converts the structure of the one or more optical vortices into output data corresponding to the structure ;
the data converter determines a set of successively varying geometric shapes, the set of images being imposed on the optical vortex and rotated to follow the direction of the vortex polarity;
The data converter determines the set of geometric shapes by measuring the relative phase change between signal bursts emitted by two orthogonal gel fiber loops, and converts vortex outputs into the output data based on this determination. 1. A gel processing method for processing a gel-like material , comprising:
preparing a fiber of gel-like material of helical nanowires that acts as a neural network to build a superstructure containing self-assembled structures in a container, the output product of the one-step synthesis being used as starting material for the next step;
generating loops of intertwined electromagnetic fields , said loops generating all loops of variables in the input data, and inducing self-assembly of a gel-like material that synthesizes a unified self-assembly architecture of the clock;
A method for processing a gel , comprising: