CN114217849B - Interpretation methods, devices and computer equipment for artificial intelligence model algorithms - Google Patents

Abstract

The present application relates to a method, apparatus, computer device, storage medium and computer program product for interpretation of artificial intelligence model algorithms. The method comprises the following steps: the method comprises the steps of obtaining electromagnetic signals of a to-be-detected group and electromagnetic signals of a comparison group of a chip, obtaining an electromagnetic field evolution diagram corresponding to an artificial intelligent model algorithm in operation according to the electromagnetic signals of the to-be-detected group and the electromagnetic signals of the comparison group, obtaining a current signal change diagram corresponding to the artificial intelligent model algorithm in operation according to the electromagnetic field evolution diagram through a Maxwell equation set, and obtaining an operation logic diagram of the artificial intelligent model algorithm according to the current signal change diagram and by combining physical logic of the chip. Based on the obtained operation logic diagram of the artificial intelligence model algorithm, the operation logic of the artificial intelligence model algorithm can be clearly explained.

Description

Interpretation methods, devices and computer equipment for artificial intelligence model algorithms

Technical field

This application relates to the field of artificial intelligence technology, and in particular to an interpretation method, device, computer equipment, storage medium and computer program product for an artificial intelligence model algorithm.

Background technique

With the development of artificial intelligence technology, artificial intelligence model algorithms are becoming increasingly complex. Artificial intelligence explainability has become an effective means to help humans understand the working principles of artificial intelligence. Explainability is the degree to which the decision-making mechanism of an artificial intelligence system can be understood by humans. It is also Artificial intelligence model algorithm is one of the most difficult, most critical and hottest research topics.

At present, the research on the interpretability of artificial intelligence model algorithms is mainly based on the method of constructing equivalent models or performing local explanations from the perspective of the software model algorithm itself. However, this method of constructing equivalent models or performing local explanations has The problem of poor interpretability of artificial intelligence model algorithms.

Contents of the invention

Based on this, it is necessary to provide an explanation method, device, computer equipment, computer readable storage medium and computer program product for the artificial intelligence model algorithm to clearly explain the artificial intelligence model in order to solve the technical problem of poor interpretability of the traditional artificial intelligence model algorithm. The operating logic of the algorithm.

In the first aspect, this application provides an explanation method for artificial intelligence model algorithms. The methods include:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group, where the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm;

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained;

According to the electromagnetic field evolution diagram and through Maxwell's equations, the corresponding current signal change diagram when the artificial intelligence model algorithm is running is obtained;

Based on the current signal change diagram and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, obtaining the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running based on the electromagnetic signals of the test group and the comparison group of electromagnetic signals includes:

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running is obtained;

According to the electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained.

In one embodiment, based on the electromagnetic signals of the group to be measured and the electromagnetic signals of the comparison group, obtaining the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running includes:

According to the electromagnetic signal of the group to be measured, the electromagnetic field intensity of the group to be measured changes with time;

According to the electromagnetic signal of the comparison group, the electromagnetic field intensity of the comparison group changing with time is obtained;

Subtract the time-varying electromagnetic field intensity of the test group from the time-varying electromagnetic field intensity of the comparison group to obtain the corresponding time-varying electromagnetic field intensity when the artificial intelligence model algorithm is running.

In one embodiment, according to the electromagnetic field evolution diagram and through Maxwell's equations, obtaining the corresponding current signal change diagram when the artificial intelligence model algorithm is running includes:

According to the electromagnetic field evolution diagram and through Maxwell's equations, the electric field intensity changing with time is obtained;

According to the electric field intensity that changes with time, the time-varying current is obtained;

According to the current changing with time, the corresponding current signal change diagram when the artificial intelligence model algorithm is running is obtained.

In one embodiment, based on the current signal change diagram and combined with the physical logic of the chip, obtaining the operation logic diagram of the artificial intelligence model algorithm includes:

According to the current signal change diagram, the initial operating logic of the artificial intelligence model algorithm is obtained;

Based on the initial operating logic and combined with the physical logic of the chip, the operating logic diagram of the artificial intelligence model algorithm is obtained.

In one embodiment, the explanation method of the artificial intelligence model algorithm further includes:

Analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm;

According to the operating logic of the artificial intelligence model algorithm, the interpretability results of the artificial intelligence model algorithm are obtained.

In a second aspect, this application also provides an explanation device for artificial intelligence model algorithms. The device includes:

The signal acquisition module is used to obtain the electromagnetic signal of the test group and the comparison group of the chip. The electromagnetic signal of the test group is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the comparison group electromagnetic signal is when the chip is not running the artificial intelligence model. Electromagnetic signals during algorithm;

The electromagnetic field evolution diagram acquisition module is used to obtain the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running based on the electromagnetic signals of the test group and the comparison group electromagnetic signals;

The current signal change diagram acquisition module is used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running based on the electromagnetic field evolution diagram and Maxwell's equations;

The operation logic diagram acquisition module is used to obtain the operation logic diagram of the artificial intelligence model algorithm based on the current signal change diagram and the physical logic of the chip. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In a third aspect, this application also provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running. According to The current signal change diagram, combined with the physical logic of the chip, obtains the operation logic diagram of the artificial intelligence model algorithm. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In a fourth aspect, this application also provides a computer-readable storage medium. The computer-readable storage medium has a computer program stored thereon, and when the computer program is executed by the processor, the following steps are implemented:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running. According to The current signal change diagram, combined with the physical logic of the chip, obtains the operation logic diagram of the artificial intelligence model algorithm. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In a fifth aspect, this application also provides a computer program product. The computer program product includes a computer program that implements the following steps when executed by a processor:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running. According to The current signal change diagram, combined with the physical logic of the chip, obtains the operation logic diagram of the artificial intelligence model algorithm. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

The interpretation method, device, computer equipment, storage medium and computer program product of the above-mentioned artificial intelligence model algorithm obtains the electromagnetic signal of the group to be tested and the electromagnetic signal of the comparison group of the chip, where the electromagnetic signal of the group to be tested is the chip running the artificial intelligence model algorithm. The electromagnetic signal at the time, the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm. According to the electromagnetic signal of the group to be tested and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution Figure, through Maxwell's equations, obtains the corresponding current signal change diagram when the artificial intelligence model algorithm is running. According to the current signal change diagram, combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. The operation logic diagram is used to explain artificial intelligence. The operating logic of the intelligent model algorithm. The above scheme obtains the electromagnetic field evolution diagram corresponding to the operation of the artificial intelligence model algorithm by obtaining the electromagnetic signal of the chip. Through Maxwell's equations, the current signal change diagram is obtained. Combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. Ability to clearly explain the operating logic of artificial intelligence model algorithms.

Description of the drawings

Figure 1 is an application environment diagram of the interpretation method of the artificial intelligence model algorithm in one embodiment;

Figure 2 is a schematic flow chart of an explanation method of an artificial intelligence model algorithm in one embodiment;

Figure 3 is a schematic flow chart of the interpretation steps of the artificial intelligence model algorithm in one embodiment;

Figure 4 is a schematic flow chart of an interpretation method of an artificial intelligence model algorithm in another embodiment;

Figure 5 is a structural block diagram of an interpretation device for an artificial intelligence model algorithm in one embodiment;

Figure 6 is an internal structure diagram of a computer device in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application and are not used to limit the present application.

The explanation method of the artificial intelligence model algorithm provided by the embodiment of the present application can be applied in the application environment as shown in Figure 1. Among them, the terminal 102 communicates with the chip 104 through the network. The data storage system may store data that terminal 102 needs to process. The data storage system can be integrated on the terminal 102, or placed on the cloud or other network servers. The terminal 102 obtains the electromagnetic signal of the test group and the comparison group electromagnetic signal of the chip 104. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip 104 is running the artificial intelligence model algorithm, and the comparison group electromagnetic signal is when the chip 104 is not running the artificial intelligence model. For the electromagnetic signal during the algorithm, based on the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the electromagnetic field evolution diagram corresponding to the operation of the artificial intelligence model algorithm is obtained. Based on the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the electromagnetic field evolution diagram corresponding to the operation of the artificial intelligence model algorithm. The current signal change diagram is based on the current signal change diagram and combined with the physical logic of the chip 104 to obtain the operation logic diagram of the artificial intelligence model algorithm. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablets, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. . Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc.

In one embodiment, as shown in Figure 2, a method for explaining an artificial intelligence model algorithm is provided. The method is explained by taking the method applied to the terminal 102 in Figure 1 as an example, and includes the following steps:

S200: Obtain the electromagnetic signal of the test group and the comparison group electromagnetic signal of the chip.

Among them, artificial intelligence is the theory, method, technology and system that can simulate, extend and expand human intelligence. The development of artificial intelligence model algorithms such as neural network models is inspired by the progress of brain neuroscience. Therefore, for the issue of interpretability of artificial intelligence model algorithms, we can also refer to the research methods of brain function mechanisms. Research on the mechanism of brain function generally uses functional magnetic resonance imaging (fMRI), electroencephalography (EEG), magnetoencephalography (MEG) and other instruments that can record blood flow changes during brain activity or brain electrical activity. While performing an experimental task, the brain activity is recorded and studied through image presentation. For example, when studying the question "whether the face recognition process is affected by emotion", an experiment will be designed that includes the task of identifying faces with different expressions. When the researchers asked subjects to judge whether the face pictures they saw belonged to a certain emotion, they also used electroencephalography to record their brain activity. Finally, they compared the accuracy, time and brain activity of the subjects' judgments. The high and low ranges are used to determine the functional area and reaction speed of the brain responsible for facial recognition and emotional judgment. Referring to human brain research methods, the artificial intelligence model algorithm is equivalent to the virtual logic network of the human brain. The chip responsible for the operation and calculation of the artificial intelligence model algorithm is equivalent to the physical entity of the human brain. Obtaining the electromagnetic signal of the chip is equivalent to recording the activities of the human brain.

Specifically, the electromagnetic signals of the test group and the comparison group of the chip are obtained, where the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the comparison group electromagnetic signal is when the chip is not running the artificial intelligence model algorithm. For electromagnetic signals, the experimental conditions of the comparison group are exactly the same as those of the test group except that the artificial intelligence model algorithm is not run. Optionally, the chips in this solution include but are not limited to ordinary chips or artificial intelligence chips. In order to obtain electromagnetic signals as accurately as possible, an electromagnetic shielding device can be set up in the chip test area to shield the electromagnetic field signal interference in the test area.

S400: Based on the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, obtain the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running.

Among them, artificial intelligence model algorithms are divided into supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning according to different model training methods. For example, supervised learning models can include linear classifier models (Linear classifier), support vector machine models (( Support Vector Machine), Naive Bayes Classifier, K-nearest neighbor, DecisionTree, Linear Regression, Regression Tree ; The unsupervised learning model mainly includes: data clustering model (K-means), data dimensionality reduction model (Principal ComponentAnalysis). According to the electromagnetic field theory of Maxwell's equation, a changing magnetic field produces an electric field, a changing electric field produces a magnetic field, and the space around the charge exists Electric field, the movement of electric charges causes the electric field to move, so the moving charges produce a magnetic field. Current is the flow of charges, so the current will produce a magnetic field. A constant magnetic field will be produced around a constant current, and an oscillating electric field will produce an oscillating electric field. According to the obtained chip operation The electromagnetic signal generated when the artificial intelligence model algorithm is used and the electromagnetic signal generated when the chip is not running the artificial intelligence algorithm, and then the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained, which is used to reflect the operating rules of the artificial intelligence model algorithm and observe its changes. The activity level determines the usage of chip resources called by the artificial intelligence model algorithm.

Specifically, based on the electromagnetic signals of the group to be measured and the electromagnetic signals of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained.

S600: Based on the electromagnetic field evolution diagram and Maxwell's equations, obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running.

Among them, when both the electric field and the magnetic field change with time, the magnetic field excited by the changing electric field and the electric field excited by the changing magnetic field are collectively called the time-varying electromagnetic field. The time-varying electromagnetic fields excite each other and have wave characteristics. The time-varying electromagnetic field obeys Maxwell's equation. Maxwell's equations include Gauss's electric field law that describes electrostatics, Gauss's magnetic field law that describes magnetostatics, Faraday's law that describes magnetism and electricity, and Ampere-Maxwell's law that describes electricity and magnetism. It can be seen from Maxwell's equations that changing magnetic fields can excite vortices. Electric field, the changing electric field can excite the vortex magnetic field. The electric field and the magnetic field are interconnected and excite each other to form a unified electromagnetic field. Through the change of the magnetic field, the change of the electric field can be calculated, and then the change diagram of the current signal can be obtained.

Specifically, according to the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running, through Maxwell's equations, the corresponding current signal change diagram when the artificial intelligence model algorithm is running can be obtained.

S800: Based on the current signal change diagram and combined with the physical logic of the chip, obtain the operation logic diagram of the artificial intelligence model algorithm.

Among them, the chip is the carrier of the integrated circuit, and the integrated circuit is a large-scale, high-performance computing module composed of multiple logical operation units. The physical logic of the chip refers to each physical logical unit of the integrated circuit in the chip. Depending on the current signal, the chip shows an operating logic of 0 or 1. Combined with the physical logic of the chip, the operating logic diagram of the artificial intelligence model algorithm can be obtained. The operating logic diagram is used to explain the operating logic of the artificial intelligence model algorithm.

Specifically, based on the current signal change diagram and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained.

In the explanation method of the above artificial intelligence model algorithm, the electromagnetic signal of the test group and the comparison group electromagnetic signal of the chip are obtained. Among them, the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip runs the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the chip. The electromagnetic signal when the artificial intelligence model algorithm is not running. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, the artificial intelligence model is obtained through Maxwell's equations. The corresponding current signal change diagram when the algorithm is running. Based on the current signal change diagram and the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm can be obtained, which can clearly explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, as shown in Figure 3, S400 obtains the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running based on the electromagnetic signals of the group to be measured and the electromagnetic signals of the comparison group, including:

S420: According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, obtain the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running.

S440: According to the electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, obtain the electromagnetic field evolution diagram corresponding to when the artificial intelligence model algorithm is running.

In this embodiment, according to the acquired electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group chip, the electromagnetic signal is converted into a function of the change of the electromagnetic field intensity with time. The corresponding electromagnetic field intensity that changes with time when the intelligent model algorithm is running. According to the electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained, which can reflect the operating rules of the artificial intelligence model algorithm.

According to the solution of the above embodiment, the electromagnetic field intensity corresponding to the time-changing electromagnetic field intensity when the artificial intelligence model algorithm is running is obtained through the electromagnetic signal of the test group and the comparison group electromagnetic signal. According to the electromagnetic field intensity corresponding to the time-varying electromagnetic field intensity when the artificial intelligence model algorithm is running, Obtaining the electromagnetic field evolution diagram corresponding to the operation of the artificial intelligence model algorithm can provide a prerequisite for supporting the explanation of the operation logic of the artificial intelligence model algorithm.

In another embodiment, according to the electromagnetic signal of the group to be measured and the electromagnetic signal of the comparison group, obtaining the electromagnetic field intensity corresponding to the time change when the artificial intelligence model algorithm is running includes: according to the electromagnetic signal of the group to be measured, obtaining the change of the group to be measured over time. The electromagnetic field intensity of the comparison group; according to the electromagnetic signal of the comparison group, obtain the electromagnetic field intensity of the comparison group that changes with time; subtract the electromagnetic field intensity that changes with time of the test group and the electromagnetic field intensity that changes with time of the comparison group to obtain the running time correspondence of the artificial intelligence model algorithm The intensity of the electromagnetic field that changes with time.

In this embodiment, the electromagnetic field intensity of the group to be measured changes with time according to the electromagnetic signal of the group to be measured; the electromagnetic field intensity of the group to be measured changes with time is obtained based on the electromagnetic signal of the comparison group. Specifically, the electromagnetic signal of the group to be measured is converted into The electromagnetic field intensity changes with time as a function to obtain the electromagnetic field intensity of the group to be tested as the electromagnetic field intensity changes with time. Convert the electromagnetic signal of the comparison group into the electromagnetic field intensity with time changes as a function to obtain the electromagnetic field intensity of the comparison group with time. Run the artificial intelligence model on the chip. The time-varying electromagnetic field intensity during the algorithm is subtracted from the time-varying electromagnetic field intensity when the chip is not running the artificial intelligence model algorithm, thereby eliminating the influence of the electromagnetic signal generated when the chip itself is running, and obtaining a corresponding response that only reflects the running time of the artificial intelligence model algorithm. The intensity of the electromagnetic field that changes with time.

According to the solution of the above embodiment, by converting the electromagnetic signal into a function of the change of the electromagnetic field intensity with time, based on the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the electromagnetic field intensity that only reflects the change with time corresponding to the running of the artificial intelligence model algorithm is obtained. It can provide a premise to support the explanation of the operating logic of artificial intelligence model algorithms.

In another embodiment, as shown in Figure 4, S600 uses Maxwell's equations based on the electromagnetic field evolution diagram to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running, including:

S620: According to the electromagnetic field evolution diagram, obtain the electric field intensity changing with time through Maxwell's equations.

S640: Obtain the time-varying current based on the time-varying electric field intensity.

S660: According to the current changing with time, obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running.

In this embodiment, according to Maxwell's electromagnetic field theory, the changing magnetic field excites the induced electric field. The induced electric field is the vortex field. According to the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running, the electric field intensity changing with time is obtained through Maxwell's equations. , according to the electric field intensity that changes with time, the positive charges move in the direction of the electric field line, and the negative charges move in the opposite direction of the electric field. The directional movement of the charges forms a current. According to the electric field intensity that changes with time, a time-varying current is obtained, and further , based on the current changing with time, a current signal change diagram is obtained, which is the corresponding current signal change diagram when the artificial intelligence model algorithm is running.

According to the solution of the above embodiment, Maxwell's equations are used to obtain the electric field intensity that changes with time according to the electromagnetic field evolution diagram. According to the electric field intensity that changes with time, the current that changes with time is obtained. According to the current that changes with time, the artificial intelligence model is obtained. The corresponding current signal change diagram when the algorithm is running can provide a prerequisite for supporting the explanation of the operation logic of the artificial intelligence model algorithm.

In another embodiment, S800 obtains the operation logic diagram of the artificial intelligence model algorithm based on the current signal change diagram and the physical logic of the chip, including: obtaining the initial operation logic of the artificial intelligence model algorithm according to the current signal change diagram; Logic, combined with the physical logic of the chip, obtains the operating logic diagram of the artificial intelligence model algorithm.

In this embodiment, the initial operation logic of the artificial intelligence model algorithm is obtained based on the corresponding current signal change diagram when the artificial intelligence model algorithm is running. For example, if the current signal is 0, it means that a certain PN junction in the chip is not conducting, and the artificial intelligence model algorithm The model algorithm is not executed here, and the current signal is 1, indicating that the PN junction is conductive, and the artificial intelligence model algorithm is executed here. According to the initial operation logic of the artificial intelligence model algorithm and the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained.

According to the solution of the above embodiment, the initial operation logic of the artificial intelligence model algorithm is obtained based on the current signal change diagram. Based on the initial operation logic and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained, which can support the interpretation of the artificial intelligence model algorithm. Run logic.

In another embodiment, the explanation of the artificial intelligence model algorithm also includes: analyzing the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm, and obtaining the interpretability result of the artificial intelligence model algorithm based on the operation logic of the artificial intelligence model algorithm.

In this embodiment, based on the obtained operating logic diagram of the artificial intelligence model algorithm, the output results and operating logic of the artificial intelligence model algorithm are analyzed, the learning reasoning and decision-making process of the algorithm is evaluated, and the interpretable results of the artificial intelligence model algorithm are formed.

According to the solution of the above embodiment, the operation logic of the artificial intelligence model algorithm is obtained by analyzing the operation logic diagram. According to the operation logic of the artificial intelligence model algorithm, the interpretability result of the artificial intelligence model algorithm is obtained, which can support the explanation of the operation of the artificial intelligence model algorithm. logical effect.

In order to explain in detail the explanation method and effect of the artificial intelligence model algorithm in this solution, the following is a most detailed example:

Use a 50 micron * 50 micron square electromagnetic probe to collect the electromagnetic signal of the chip. The probe array is covered in parallel on the running chip. The height of the probe array from the chip surface remains unchanged. An electromagnetic shielding device is installed in the chip area corresponding to the electromagnetic probe. , set up the test group and the comparison group respectively. Among them, the test group tests the electromagnetic signal when the chip runs the artificial intelligence model algorithm, and the comparison group tests the electromagnetic signal when the chip does not run the artificial intelligence model algorithm. The comparison group does not run the artificial intelligence model. Except for the algorithm, other conditions are consistent with the test group. Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running. According to The current signal change diagram, combined with the physical logic of the chip, obtains the operation logic diagram of the artificial intelligence model algorithm. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm. According to the electromagnetic signal of the group to be tested, the electromagnetic field intensity of the group to be tested is obtained as time changes. According to the electromagnetic signal of the comparison group, the electromagnetic field intensity of the comparison group is obtained as time changes. Subtract the electromagnetic field intensity to obtain the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running. According to the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, obtain the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running. . According to the electromagnetic field evolution diagram and through Maxwell's equations, the electric field intensity that changes with time is obtained. According to the electric field intensity that changes with time, the current that changes with time is obtained. According to the current that changes with time, the corresponding current when the artificial intelligence model algorithm is running is obtained. Signal change diagram: According to the current signal change diagram, the initial operation logic of the artificial intelligence model algorithm is obtained. Based on the initial operation logic, combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. Analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm. According to the operation logic of the artificial intelligence model algorithm, obtain the interpretability results of the artificial intelligence model algorithm.

The solution of the above embodiment is to obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group. The electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running artificial intelligence. For the electromagnetic signal during model algorithm, based on the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the electromagnetic field evolution diagram corresponding to the running time of the artificial intelligence model algorithm is obtained. According to the electromagnetic field evolution diagram, Maxwell's equations are used to obtain the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running. According to the current signal change diagram, combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm can be obtained, which can clearly explain the operation logic of the artificial intelligence model algorithm.

It should be understood that although the steps in the flowcharts involved in the above-mentioned embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated in this article, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above embodiments may include multiple steps or stages. These steps or stages are not necessarily executed at the same time, but may be completed at different times. The execution order of these steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

Based on the same inventive concept, embodiments of the present application also provide an artificial intelligence model algorithm interpretation device for implementing the above-mentioned artificial intelligence model algorithm interpretation method. The implementation scheme for solving the problem provided by this device is similar to the implementation scheme recorded in the above method. Therefore, the specific limitations in the device embodiments for explaining one or more artificial intelligence model algorithms provided below can be found in the above article on artificial intelligence. The limitations of the interpretation method of the model algorithm will not be repeated here.

In one embodiment, as shown in Figure 5, an artificial intelligence model algorithm interpretation device 500 is provided, including: a signal acquisition module 520, an electromagnetic field evolution diagram acquisition module 540, a current signal change diagram acquisition module 560 and an operation logic diagram. Get module 580, which:

The signal acquisition module 520 is used to obtain the electromagnetic signals of the test group and the comparison group of electromagnetic signals of the chip. The electromagnetic signals of the group to be tested are the electromagnetic signals when the chip is running the artificial intelligence model algorithm, and the comparison group electromagnetic signals are when the chip is not running artificial intelligence. Electromagnetic signals during model algorithm;

The electromagnetic field evolution diagram acquisition module 540 is used to obtain the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running based on the electromagnetic signals of the group to be measured and the electromagnetic signals of the comparison group;

The current signal change diagram acquisition module 560 is used to obtain the corresponding current signal change diagram when the artificial intelligence model algorithm is running based on the electromagnetic field evolution diagram and Maxwell's equations;

The operation logic diagram acquisition module 580 is used to obtain the operation logic diagram of the artificial intelligence model algorithm based on the current signal change diagram and the physical logic of the chip. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In the device for interpreting the above-mentioned artificial intelligence model algorithm, the electromagnetic signal of the group to be tested and the electromagnetic signal of the comparison group are obtained by The electromagnetic signal when the artificial intelligence model algorithm is not running. According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained. According to the electromagnetic field evolution diagram, the artificial intelligence model is obtained through Maxwell's equations. The corresponding current signal change diagram when the algorithm is running. Based on the current signal change diagram and the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm can be obtained, which can clearly explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, the electromagnetic field evolution diagram acquisition module 540 is also used to obtain the electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running based on the electromagnetic signal of the group to be measured and the electromagnetic signal of the comparison group; according to the running time of the artificial intelligence model algorithm The corresponding electromagnetic field intensity changes with time, and the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained.

In one embodiment, the electromagnetic field evolution diagram acquisition module 540 is also used to obtain the electromagnetic field intensity of the group to be measured that changes with time based on the electromagnetic signal of the group to be measured; and based on the electromagnetic signal of the comparison group, obtain the electromagnetic field intensity that changes with time of the comparison group; The time-varying electromagnetic field intensity of the test group is subtracted from the time-varying electromagnetic field intensity of the comparison group to obtain the corresponding time-varying electromagnetic field intensity when the artificial intelligence model algorithm is running.

In one embodiment, the current signal change diagram acquisition module 560 is also used to obtain the electric field intensity changing with time according to the electromagnetic field evolution diagram through Maxwell's equations; according to the electric field intensity changing with time, obtain the current changing with time; according to The current changes with time, and the corresponding current signal change diagram is obtained when the artificial intelligence model algorithm is running.

In one embodiment, the operation logic diagram acquisition module 580 is also used to obtain the initial operation logic of the artificial intelligence model algorithm based on the current signal change diagram; and obtain the operation logic of the artificial intelligence model algorithm based on the initial operation logic, combined with the physical logic of the chip. picture.

In one embodiment, the artificial intelligence model algorithm interpretation device 500 is also used to analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm; and obtain the interpretability results of the artificial intelligence model algorithm based on the operation logic of the artificial intelligence model algorithm.

Each module in the interpretation device of the above-mentioned artificial intelligence model algorithm can be implemented in whole or in part by software, hardware, and combinations thereof. Each of the above modules may be embedded in or independent of the processor of the computer device in the form of hardware, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 6 . The computer device includes a processor, memory, and network interfaces connected through a system bus. Wherein, the processor of the computer device is used to provide computing and control capabilities.

The memory of the computer device includes non-volatile storage media and internal memory. The non-volatile storage medium stores operating systems, computer programs and databases. This internal memory provides an environment for the execution of operating systems and computer programs in non-volatile storage media. The database of the computer equipment is used to store the corresponding electromagnetic field evolution diagram, current signal change diagram and operation logic diagram when the artificial intelligence model algorithm is running. The network interface of the computer device is used to communicate with external terminals through a network connection. The computer program, when executed by the processor, implements an interpretation method of an artificial intelligence model algorithm.

Those skilled in the art can understand that the structure shown in Figure 6 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Specific computer equipment can May include more or fewer parts than shown, or combine certain parts, or have a different arrangement of parts.

In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, it implements the following steps:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group, where the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm;

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained;

According to the electromagnetic field evolution diagram and through Maxwell's equations, the corresponding current signal change diagram when the artificial intelligence model algorithm is running is obtained;

Based on the current signal change diagram and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, the processor also implements the following steps when executing the computer program:

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running is obtained. According to the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, the artificial intelligence model algorithm operation is obtained. The electromagnetic field evolution diagram corresponding to time.

In one embodiment, the processor also implements the following steps when executing the computer program:

According to the electromagnetic signal of the group to be tested, the electromagnetic field intensity of the group to be tested is obtained as time changes. According to the electromagnetic signal of the comparison group, the electromagnetic field intensity of the comparison group is obtained as time changes. The electromagnetic field intensity is subtracted to obtain the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running.

In one embodiment, the processor also implements the following steps when executing the computer program:

According to the electromagnetic field evolution diagram and through Maxwell's equations, the electric field intensity that changes with time is obtained. According to the electric field intensity that changes with time, the current that changes with time is obtained. According to the current that changes with time, the corresponding current when the artificial intelligence model algorithm is running is obtained. Signal change graph.

In one embodiment, the processor also implements the following steps when executing the computer program:

According to the current signal change diagram, the initial operating logic of the artificial intelligence model algorithm is obtained. Based on the initial operating logic and combined with the physical logic of the chip, the operating logic diagram of the artificial intelligence model algorithm is obtained.

In one embodiment, the processor also implements the following steps when executing the computer program:

Analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm. According to the operation logic of the artificial intelligence model algorithm, obtain the interpretability results of the artificial intelligence model algorithm.

In one embodiment, a computer-readable storage medium is provided with a computer program stored thereon. When the computer program is executed by a processor, the following steps are implemented:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group, where the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm;

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained;

According to the electromagnetic field evolution diagram and through Maxwell's equations, the corresponding current signal change diagram when the artificial intelligence model algorithm is running is obtained;

Based on the current signal change diagram and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running is obtained. According to the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, the artificial intelligence model algorithm operation is obtained. The electromagnetic field evolution diagram corresponding to time.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic signal of the group to be tested, the electromagnetic field intensity of the group to be tested is obtained as time changes. According to the electromagnetic signal of the comparison group, the electromagnetic field intensity of the comparison group is obtained as time changes. The electromagnetic field intensity is subtracted to obtain the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic field evolution diagram and through Maxwell's equations, the electric field intensity that changes with time is obtained. According to the electric field intensity that changes with time, the current that changes with time is obtained. According to the current that changes with time, the corresponding current when the artificial intelligence model algorithm is running is obtained. Signal change graph.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the current signal change diagram, the initial operating logic of the artificial intelligence model algorithm is obtained. Based on the initial operating logic and combined with the physical logic of the chip, the operating logic diagram of the artificial intelligence model algorithm is obtained.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

Analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm. According to the operation logic of the artificial intelligence model algorithm, obtain the interpretability results of the artificial intelligence model algorithm.

In one embodiment, a computer program product is provided, comprising a computer program that when executed by a processor implements the following steps:

Obtain the electromagnetic signal of the chip to be tested and the electromagnetic signal of the comparison group, where the electromagnetic signal of the group to be tested is the electromagnetic signal when the chip is running the artificial intelligence model algorithm, and the electromagnetic signal of the comparison group is the electromagnetic signal when the chip is not running the artificial intelligence model algorithm;

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field evolution diagram when the artificial intelligence model algorithm is running is obtained;

According to the electromagnetic field evolution diagram and through Maxwell's equations, the corresponding current signal change diagram when the artificial intelligence model algorithm is running is obtained;

Based on the current signal change diagram and combined with the physical logic of the chip, the operation logic diagram of the artificial intelligence model algorithm is obtained. The operation logic diagram is used to explain the operation logic of the artificial intelligence model algorithm.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic signal of the test group and the electromagnetic signal of the comparison group, the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running is obtained. According to the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running, the artificial intelligence model algorithm operation is obtained. The electromagnetic field evolution diagram corresponding to time.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic signal of the group to be tested, the electromagnetic field intensity of the group to be tested is obtained as time changes. According to the electromagnetic signal of the comparison group, the electromagnetic field intensity of the comparison group is obtained as time changes. The electromagnetic field intensity is subtracted to obtain the corresponding electromagnetic field intensity that changes with time when the artificial intelligence model algorithm is running.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the electromagnetic field evolution diagram and through Maxwell's equations, the electric field intensity that changes with time is obtained. According to the electric field intensity that changes with time, the current that changes with time is obtained. According to the current that changes with time, the corresponding current when the artificial intelligence model algorithm is running is obtained. Signal change graph.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

According to the current signal change diagram, the initial operating logic of the artificial intelligence model algorithm is obtained. Based on the initial operating logic and combined with the physical logic of the chip, the operating logic diagram of the artificial intelligence model algorithm is obtained.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

Analyze the operation logic diagram to obtain the operation logic of the artificial intelligence model algorithm. According to the operation logic of the artificial intelligence model algorithm, obtain the interpretability results of the artificial intelligence model algorithm.

It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all It is information and data authorized by the user or fully authorized by all parties.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage. In the media, when executed, the computer program may include the processes of the above method embodiments. Any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetic variable memory (Magnetoresistive Random) Access Memory (MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (PCM), graphene memory, etc. Volatile memory may include random access memory (Random Access Memory, RAM) or external cache memory. By way of illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM). The databases involved in the various embodiments provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include blockchain-based distributed databases, etc., but are not limited thereto. The processors involved in the various embodiments provided in this application may be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to this.

The technical features of the above embodiments can be combined in any way. To simplify the description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, all possible combinations should be used. It is considered to be within the scope of this manual.

The above-described embodiments only express several implementation modes of the present application, and their descriptions are relatively specific and detailed, but should not be construed as limiting the patent scope of the present application. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all fall within the protection scope of the present application. Therefore, the scope of protection of this application should be determined by the appended claims.

Claims (7)

1. A method of interpretation of an artificial intelligence model algorithm, the method comprising:

acquiring electromagnetic signals to be tested of a chip and electromagnetic signals of a comparison group, wherein the electromagnetic signals to be tested are electromagnetic signals when the chip runs an artificial intelligent model algorithm, and the electromagnetic signals of the comparison group are electromagnetic signals when the chip does not run the artificial intelligent model algorithm;

Obtaining the electromagnetic field intensity of the group to be tested, which changes along with time, according to the electromagnetic signals of the group to be tested; obtaining the electromagnetic field intensity of the contrast group changing along with time according to the electromagnetic signals of the contrast group; subtracting the time-varying electromagnetic field intensity of the to-be-detected group from the time-varying electromagnetic field intensity of the comparison group to obtain the corresponding time-varying electromagnetic field intensity when the artificial intelligent model algorithm operates; obtaining an electromagnetic field evolution diagram corresponding to the artificial intelligence model algorithm according to the electromagnetic field intensity corresponding to the artificial intelligence model algorithm during operation and changing along with time;

according to the electromagnetic field evolution diagram, obtaining a corresponding current signal variation diagram when an artificial intelligent model algorithm operates through a Maxwell equation set;

and according to the current signal change diagram, combining the physical logic of the chip to obtain an operation logic diagram of the artificial intelligent model algorithm, wherein the operation logic diagram is used for explaining the operation logic of the artificial intelligent model algorithm.

2. The interpretation method of artificial intelligence model algorithm as claimed in claim 1, wherein obtaining a current signal variation map corresponding to the operation of artificial intelligence model algorithm through maxwell's equations based on the electromagnetic field evolution map comprises:

According to the electromagnetic field evolution diagram, obtaining the electric field intensity changing along with time through a Maxwell equation set;

obtaining a current which changes with time according to the electric field intensity which changes with time;

and obtaining a corresponding current signal change diagram when the artificial intelligent model algorithm operates according to the current changing along with time.

3. The method of claim 2, wherein obtaining an operational logic map of an artificial intelligence model algorithm based on the current signal variation map in combination with physical logic of the chip comprises:

obtaining initial operation logic of an artificial intelligent model algorithm according to the current signal change diagram;

and according to the initial operation logic, combining the physical logic of the chip to obtain an operation logic diagram of an artificial intelligent model algorithm.

4. The method of interpretation of an artificial intelligence model algorithm of claim 1, further comprising:

analyzing the operation logic diagram to obtain operation logic of an artificial intelligent model algorithm;

and obtaining an interpretive result of the artificial intelligent model algorithm according to the operation logic of the artificial intelligent model algorithm.

5. An interpretation apparatus for an artificial intelligence model algorithm, the apparatus comprising:

the signal acquisition module is used for acquiring electromagnetic signals of a group to be detected and electromagnetic signals of a comparison group of chips, wherein the electromagnetic signals of the group to be detected are electromagnetic signals when the chips run an artificial intelligent model algorithm, and the electromagnetic signals of the comparison group are electromagnetic signals when the chips do not run the artificial intelligent model algorithm;

the electromagnetic field evolution diagram acquisition module is used for acquiring the electromagnetic field intensity of the to-be-measured group changing along with time according to the electromagnetic signals of the to-be-measured group; obtaining the electromagnetic field intensity of the contrast group changing along with time according to the electromagnetic signals of the contrast group; subtracting the time-varying electromagnetic field intensity of the to-be-detected group from the time-varying electromagnetic field intensity of the comparison group to obtain the corresponding time-varying electromagnetic field intensity when the artificial intelligent model algorithm operates; obtaining an electromagnetic field evolution diagram corresponding to the artificial intelligence model algorithm according to the electromagnetic field intensity corresponding to the artificial intelligence model algorithm during operation and changing along with time;

the current signal change diagram acquisition module is used for acquiring a corresponding current signal change diagram when the artificial intelligent model algorithm operates through a Maxwell equation set according to the electromagnetic field evolution diagram;

And the operation logic diagram acquisition module is used for acquiring an operation logic diagram of the artificial intelligent model algorithm according to the current signal change diagram and combining the physical logic of the chip, and the operation logic diagram is used for explaining the operation logic of the artificial intelligent model algorithm.

6. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 4 when the computer program is executed.

7. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 4.