Auxiliary method and system for enhancing large model forgetting through parameter extrapolation

Abstract

The invention belongs to the technical field of large language model forgetting in the field of artificial intelligence. The invention provides an auxiliary method and system for enhancing large model forgetting through parameter extrapolation. The method comprises the following steps: performing forgetting operation on a target forgetting data set through gradient rising; after the forgetting operation is finished, selecting a model parameter with optimal balance between the forgetting effect and the model effectiveness in a plurality of training rounds of the forgetting operation, and calculating a parameter updating vector in the forgetting process; according to the balanced optimal model parameter and the parameter updating vector, amplifying the parameter updating vector through parameter extrapolation so as to expand the forgetting effect to knowledge related to the target forgetting data set; according to the method, the forgetting effect of the model on related knowledge when forgetting target data is amplified by using a parameter extrapolation technology, so that more thorough forgetting is realized, the forgetting effect can be remarkably improved without additional training data, and better balance between forgetting quality and model effectiveness is realized.

Description

Technical Field

[0001] The present invention relates to the field of large language model forgetting technology in the field of artificial intelligence, and in particular to an auxiliary method and system for enhancing large model forgetting through parameter extrapolation. Background Art

[0002] The statements in this section merely provide background art related to the present invention and do not necessarily constitute prior art.

[0003] Large Language Models (LLMs) have made significant progress in the field of artificial intelligence in recent years, demonstrating powerful language generation and understanding capabilities. These models can generate high-quality text content by training on massive data and are widely used in natural language processing tasks such as machine translation, text generation, and question-answering systems. However, with the widespread application of LLMs, their training data inevitably contains harmful information, such as privacy data, copyrighted content, and inherent biases. The presence of this information may not only reduce the performance of the model, but may also cause social and legal problems. In order to solve the above problems, LLMs forgetting technology came into being. The goal of forgetting technology is to reduce or eliminate the impact of the model on specific data (such as privacy data) while trying to maintain the overall performance of the model.

[0004] The LLMs forgetting technique can be achieved by retraining the model and excluding the specified forgotten data. However, this method is computationally expensive and requires access to the original training data, which has many limitations in practical applications. Therefore, one of the current mainstream techniques is to approximate the forgetting effect through gradient ascent. These methods attempt to reverse the model's learning process on these data by maximizing the prediction loss on the target forgotten data set. Despite this, the existing gradient ascent-based forgetting methods still have the problem of poor forgetting effect. Although the gradient ascent-based methods can forget the target data to a certain extent, they often ignore the knowledge related to the logic of the target data. This related knowledge may reconstruct the forgotten target information through the model's reasoning mechanism, thereby reducing the forgetting effect. Summary of the invention

[0005] In order to address the shortcomings of the prior art, the present invention provides an auxiliary method and system for enhancing the forgetting of large models through parameter extrapolation. The parameter extrapolation technology is used to amplify the forgetting effect of the model on related knowledge when forgetting the target data, thereby achieving more thorough forgetting. The forgetting effect can be significantly improved without the need for additional training data, and a better balance is achieved between the quality of forgetting and the utility of the model.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides an auxiliary method for enhancing forgetting of large models through parameter extrapolation.

[0007] An auxiliary method for enhancing forgetting of large models through parameter extrapolation, including the following process: Perform a forget operation on the target forget data set through gradient ascent; After the forgetting operation is completed, in multiple training rounds of the forgetting operation, the model parameters that best balance the forgetting effect and model utility are selected, and the parameter update vector in the forgetting process is calculated; According to the optimally balanced model parameters and the parameter update vector, the parameter update vector is amplified by parameter extrapolation to extend the forgetting effect to knowledge related to the target forgotten data set.

[0008] In a second aspect, the present invention provides an auxiliary system for enhancing forgetting of large models through parameter extrapolation.

[0009] An auxiliary system for enhancing forgetting in large models via parameter extrapolation, including: The forgetting operation unit is configured to: perform a forgetting operation on a target forgetting data set by gradient ascent; A parameter update vector calculation unit is configured to: after the forgetting operation is completed, in multiple training rounds of the forgetting operation, select the model parameters that best balance the forgetting effect and the model utility, and calculate the parameter update vector in the forgetting process; The parameter extrapolation unit is configured to: amplify the parameter update vector by parameter extrapolation based on the optimally balanced model parameters and the parameter update vector, so as to extend the forgetting effect to the knowledge related to the target forgotten data set.

[0010] In a third aspect, the present invention provides a computer device, comprising: a processor and a computer-readable storage medium; a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in the first aspect of the present invention is implemented.

[0011] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor and executing the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in the first aspect of the present invention.

[0012] In a fifth aspect, the present invention provides a computer program product, comprising a computer program, which, when executed by a processor, implements the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in the first aspect of the present invention.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention innovatively proposes an auxiliary method for enhancing the forgetting of large models through parameter extrapolation. By using parameter extrapolation technology, the forgetting effect of the model on related knowledge when forgetting the target data is amplified, thereby achieving more thorough forgetting. The forgetting effect can be significantly improved without the need for additional training data, and a better balance is achieved between the quality of forgetting and the utility of the model.

[0014] 2. The present invention innovatively proposes a method for enhancing the forgetting effect through parameter extrapolation, aiming to significantly improve the effect of LLMs when forgetting target data and its related knowledge. Existing forgetting methods (gradient ascent) mainly focus on directly forgetting target data, while ignoring knowledge logically related to the target data. These related knowledge may reconstruct the forgotten target information through the model's reasoning mechanism, resulting in incomplete forgetting effect. The present invention uses parameter extrapolation technology to amplify the model's forgetting effect on related knowledge when forgetting target data, which can effectively avoid reconstructing forgotten information through reasoning mechanism, thereby achieving more thorough forgetting.

[0015] 3. Existing forgetting methods usually require additional training data to assist the forgetting process, which not only increases the complexity of the forgetting process, but may also introduce new privacy and security risks. The present invention operates directly based on model parameters without the need for additional training data, thereby reducing the complexity and potential risks of the forgetting process and improving the efficiency and security of the forgetting process.

[0016] 4. In practical applications, large language models may contain a large amount of sensitive information, such as personal privacy, copyrighted content, etc. The present invention can effectively reduce the risk of model leakage of sensitive information through a more thorough forgetting effect, and improve the privacy protection and security of the model. This is of great significance for protecting user privacy, complying with data protection regulations, and reducing the potential social risks of the model.

[0017] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0019] Figure 1 A schematic diagram of the utility of the model provided in Example 1 of the present invention (a higher score indicates a better utility); Figure 2 A schematic diagram of forgetting quality (a lower score indicates a better forgetting effect) provided in Example 1 of the present invention; Figure 3 A schematic flow chart of an auxiliary method for enhancing forgetting of a large model by parameter extrapolation provided in Example 1 of the present invention; Figure 4 A schematic diagram of an auxiliary system for enhancing large model forgetting through parameter extrapolation provided by Embodiment 2 of the present invention; Figure 5 A schematic diagram of a computer device provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0021] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0022] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0023] Embodiment 1: This implementation proposes an auxiliary method for enhancing the forgetting of large models through parameter extrapolation. Compared with the existing technology, it does not require additional training data and can achieve a more thorough forgetting effect. It has wide applicability and promotion value, and improves the application effect of LLMs in scenarios such as privacy protection, copyright compliance, and reducing the spread of harmful information.

[0024] Specifically in the present invention, a preliminary experiment was conducted using a virtual character dataset that contains a target forgetting set and a related knowledge set. The results show that when a model is trained on these two sets, forgetting only the target forgetting set is not enough to completely remove the knowledge. However, when related knowledge is included in the forgetting process, the forgetting effect of the model on the target forgetting set is significantly improved. These findings indicate that large language models can reconstruct target knowledge that should be forgotten through related information. Based on the above experiments, the present invention proposes an auxiliary method for enhancing forgetting of large models through parameter extrapolation, which is a plug-and-play method for auxiliary forgetting learning. The overall architecture includes the following three key processes: (1) Basic forgetting: forgetting the target data through the gradient ascent method; (2) Parameter update vector calculation: calculating the parameter update vector in the forgetting process; (3) Parameter extrapolation: enhancing the forgetting effect of related knowledge by amplifying the parameter update vector.

[0025] More specifically, the preliminary experiments of the present invention are intended to verify whether LLMs can infer the forgotten target information through relevant knowledge, thereby affecting the forgetting effect, more specifically, including: S1.1: Dataset construction.

[0026] A synthetic dataset containing a target forgetting dataset and a related knowledge dataset is constructed. The specific construction method of the dataset is as follows: Target forgetting dataset: contains specific knowledge instances that need to be forgotten, such as "patient X suffers from diabetes"; Related knowledge dataset: contains knowledge instances that are logically related to the target forgotten data, such as "someone needs to inject insulin regularly" and "insulin is the standard treatment for diabetes". These knowledge instances can be used to infer the target forgotten data through the model's reasoning mechanism; The dataset consists of 12 virtual individuals, each of which contains 10 specific attributes. Each attribute generates two question-answer pairs (K1 and K2). The K1 question-answer pair directly describes the target forgotten data, while the K2 question-answer pair contains relevant knowledge and the content of K1 can be derived through common sense reasoning.

[0027] S1.2: Model selection.

[0028] The experiment uses the LLaMA-2-7b-chat model as the base model, which performs well in natural language processing tasks and is suitable for verifying the effect of the forgetting technique.

[0029] S1.3: Experimental setup.

[0030] The experiment was divided into three situations to systematically evaluate the impact of relevant knowledge on the forgetting effect: Model : Fine-tune on a joint dataset containing the target forgotten dataset and the related knowledge dataset, and then apply the gradient ascent forgetting method only on the target forgotten dataset; Model : Fine-tune only on the target forgotten dataset and then apply the gradient ascent forgetting method; Model : Fine-tune on a joint dataset containing a target forgotten dataset and a related knowledge dataset, and then apply the forgetting method to both the target forgotten dataset and the related knowledge dataset simultaneously.

[0031] S1.4: Evaluation metrics.

[0032] The experiment uses the following two indicators to evaluate the forgetting effect: Forget Quality: The forgetting effect is measured by calculating the ROUGE-L score of the model on the target forgetting dataset. A lower ROUGE-L score indicates a better forgetting effect, such as Figure 2 As shown; Model Utility: The overall performance of the model is measured by calculating the ROUGE-L score of the model on the TruthfulQA dataset. A higher ROUGE-L score indicates better model utility. Figure 1 shown.

[0033] S1.5: Experimental results.

[0034] like Figure 1 and Figure 2 As shown, Figure 1 The ordinate in is the ROUGE-L score on the TruthfulQA dataset. Figure 2 The ordinate in is the ROUGE-L score on the target forgotten dataset. It can be seen that the model proposed in this invention can reconstruct the forgotten knowledge by using relevant knowledge. compared to, The key difference between these models is that Trained on the target forget set and the relevant knowledge set, Training is performed only on the target forgotten set. Therefore, even after forgetting the target forgotten set, The forgotten knowledge can still be reconstructed by using related knowledge, resulting in poor forgetting performance. This finding verifies that related knowledge enables LLMs to infer the forgotten information, reducing the effectiveness of forgetting. Forgetting related knowledge can improve the quality of forgetting on the target forgetting set, which is consistent with compared to, Forgetting is performed on both the target forgetting set and the related knowledge set, and the quality of forgetting on the target forgetting set is significantly improved while maintaining comparable model utility.

[0035] like Figure 3 As shown, the auxiliary method for enhancing the forgetting of a large model by parameter extrapolation provided by the present invention specifically includes: S2.1: Forgetting stage.

[0036] The target forgetting data set is forgotten by gradient ascent technique. represents a model that has only been fine-tuned without forget training, Represents a model that has only been trained with gradient ascent and forgetfulness. For any sample in the target forgetting set (The corresponding sample in the relevant knowledge set is ), in the model Gradient ascent is performed on , and the parameter update is expressed as: (1); in, is the learning rate, is the objective function of gradient ascent, In the model The gradient of Indicates that the initial parameters Next, given input Output The conditional probability of .

[0037] S2.2: Parameter update vector calculation.

[0038] After the forgetting phase, select the model parameters that best balance the forgetting effect and model utility in multiple training rounds. , calculate the parameter update vector during the forgetting process : (2); in, are the model parameters for gradient ascent and forget training, are the corresponding model parameters that have only been fine-tuned without forgetting training, The vector represents the direction and magnitude of change of the model parameters when forgetting the target data.

[0039] S2.3: Parameter extrapolation.

[0040] In order to enhance the forgetting effect of relevant knowledge without the need for additional data training, the parameter update vector is amplified by parameter extrapolation technology. , expressed as: (3); in, is the parameter update vector amplified by the parameter extrapolation technique, Used to control the amplification factor of the parameter update vector. This coefficient is a hyperparameter that can be adjusted according to the forgetting effect and model utility. ,The present invention can extend the forgetting effect to the knowledge related to the target data, thereby achieving more thorough forgetting.

[0041] The present invention significantly improves the effect of LLMs when forgetting target data and related knowledge through the design of the above steps S2.1 to S2.3. Through parameter extrapolation technology, the forgetting effect of the model on related knowledge when forgetting the target data is amplified, which can effectively avoid reconstructing forgotten information through reasoning mechanism, thereby achieving more thorough forgetting. It operates directly based on model parameters without the need for additional training data, reduces the complexity and potential risks of the forgetting process, and improves the efficiency and security of the forgetting process. Through a more thorough forgetting effect, it can effectively reduce the risk of the model leaking sensitive information and improve the privacy protection and security of the model, which is of great significance for protecting user privacy, complying with data protection regulations, and reducing the potential social risks of the model.

[0042] Embodiment 2: like Figure 4 As shown, this implementation provides an auxiliary system for enhancing large model forgetting through parameter extrapolation, including: The forgetting operation unit is configured to: perform a forgetting operation on a target forgetting data set by gradient ascent; A parameter update vector calculation unit is configured to: after the forgetting operation is completed, in multiple training rounds of the forgetting operation, select the model parameters that best balance the forgetting effect and the model utility, and calculate the parameter update vector in the forgetting process; The parameter extrapolation unit is configured to: amplify the parameter update vector by parameter extrapolation based on the optimally balanced model parameters and the parameter update vector, so as to extend the forgetting effect to the knowledge related to the target forgotten data set.

[0043] The specific working methods of the above-mentioned units are described in Example 1 and will not be repeated here.

[0044] It is understandable that the above-mentioned units can be separately or completely combined into one or several other units to constitute, or one (some) of the units can be further divided into multiple smaller units in function to constitute, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above-mentioned units are divided based on logical functions. In practical applications, the functions of one unit can also be implemented by multiple units, or the functions of multiple units can be implemented by one unit. In other embodiments of the present application, the system may also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented by the collaboration of multiple units.

[0045] According to another embodiment of the present application, the system described in this embodiment can be constructed, and the method of Example 1 of the present application can be implemented by running a computer program (including program code) capable of executing the steps involved in the corresponding method described in Example 1 on a general-purpose computing device such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). The computer program can be recorded on, for example, a computer-readable recording medium, and loaded into the above-mentioned computing device through the computer-readable recording medium and run therein.

[0046] Embodiment 3: like Figure 5 As shown, this implementation provides an electronic device, which includes a processor 1001, a communication interface 1002, and a computer-readable storage medium 1003. The processor 1001, the communication interface 1002, and the computer-readable storage medium 1003 may be connected via a bus or other means.

[0047] Among them, the communication interface 1002 is used to receive and send data, the computer-readable storage medium 1003 can be stored in the memory of the electronic device, the computer-readable storage medium 1003 is used to store a computer program, the computer program includes program instructions, and the processor 1001 is used to execute the program instructions stored in the computer-readable storage medium 1003.

[0048] The processor 1001 (or CPU (Central Processing Unit)) is the computing core and control core of the electronic device, which is suitable for implementing one or more instructions, and specifically suitable for loading and executing one or more instructions to implement corresponding method processes or corresponding functions.

[0049] The processor 1001 is configured to execute the following process: Perform a forget operation on the target forget data set through gradient ascent; After the forgetting operation is completed, in multiple training rounds of the forgetting operation, the model parameters that best balance the forgetting effect and model utility are selected, and the parameter update vector in the forgetting process is calculated; According to the optimally balanced model parameters and the parameter update vector, the parameter update vector is amplified by parameter extrapolation to extend the forgetting effect to knowledge related to the target forgotten data set.

[0050] The specific working method is described in Example 1 and will not be repeated here.

[0051] Embodiment 4: This implementation provides a computer-readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the electronic device and, of course, extended storage media supported by the electronic device. The computer-readable storage medium provides a storage space that stores the processing system of the electronic device.

[0052] In addition, the storage space also stores one or more instructions suitable for being loaded and executed by the processor, and these instructions may be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here may be a high-speed RAM memory, or a non-volatile memory, such as at least one disk memory; optionally, it may also be at least one computer-readable storage medium located away from the aforementioned processor.

[0053] In one embodiment, the computer-readable storage medium stores one or more instructions; the processor loads and executes the one or more instructions stored in the computer-readable storage medium to implement the following process: Perform a forget operation on the target forget data set through gradient ascent; After the forgetting operation is completed, in multiple training rounds of the forgetting operation, the model parameters that best balance the forgetting effect and model utility are selected, and the parameter update vector in the forgetting process is calculated; According to the optimally balanced model parameters and the parameter update vector, the parameter update vector is amplified by parameter extrapolation to extend the forgetting effect to knowledge related to the target forgotten data set.

[0054] The specific working method is described in Example 1 and will not be repeated here.

[0055] Embodiment 5: The present implementation provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device performs the following process: Perform a forget operation on the target forget data set through gradient ascent; After the forgetting operation is completed, in multiple training rounds of the forgetting operation, the model parameters that best balance the forgetting effect and model utility are selected, and the parameter update vector in the forgetting process is calculated; According to the optimally balanced model parameters and the parameter update vector, the parameter update vector is amplified by parameter extrapolation to extend the forgetting effect to knowledge related to the target forgotten data set.

[0056] The specific working method is described in Example 1 and will not be repeated here.

[0057] A person skilled in the art can appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0058] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by the computer or a data processing device such as a server, a data center, etc. that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

[0059] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An auxiliary method for enhancing forgetting of large models by parameter extrapolation, characterized in that The process includes: Perform a forget operation on the target forget data set through gradient ascent; After the forgetting operation is completed, in multiple training rounds of the forgetting operation, the model parameters that best balance the forgetting effect and model utility are selected, and the parameter update vector in the forgetting process is calculated; According to the optimally balanced model parameters and the parameter update vector, the parameter update vector is amplified by parameter extrapolation to extend the forgetting effect to knowledge related to the target forgotten data set.

2. The auxiliary method for enhancing forgetting of a large model by parameter extrapolation as claimed in claim 1, characterized in that: The target forgetting data set is forgotten by gradient ascent, including: assumed represents a model that has only been fine-tuned without forget training, Represents a model that has only been trained with gradient ascent and forgetfulness. For any sample in the target forgetting set , the parameter update of gradient ascent is expressed as: ； in, for The model parameters, is the learning rate, is the objective function of gradient ascent, In the model The gradient of For the updated Parameters, Indicates that the initial parameters Next, given input Output The conditional probability of .

3. The auxiliary method for enhancing large model forgetting by parameter extrapolation as claimed in claim 1, characterized in that: The parameter update vector represents the direction and magnitude of change of the model parameters when forgetting the target data.

4. The auxiliary method for enhancing forgetting of a large model by parameter extrapolation according to any one of claims 1 to 3, characterized in that: According to the optimal model parameters and the parameter update vector , amplifying the parameter update vector by parameter extrapolation, including: ,in, Used to control the magnification of the parameter update vector, is the parameter update vector after parameter extrapolation and amplification.

5. An auxiliary system for enhancing forgetting of large models by parameter extrapolation, characterized in that include: The forgetting operation unit is configured to: perform a forgetting operation on a target forgetting data set by gradient ascent; A parameter update vector calculation unit is configured to: after the forgetting operation is completed, in multiple training rounds of the forgetting operation, select the model parameters that best balance the forgetting effect and the model utility, and calculate the parameter update vector in the forgetting process; The parameter extrapolation unit is configured to: amplify the parameter update vector by parameter extrapolation based on the optimally balanced model parameters and the parameter update vector, so as to extend the forgetting effect to the knowledge related to the target forgotten data set.

6. The auxiliary system for enhancing large model forgetting by parameter extrapolation as claimed in claim 5, characterized in that: In the forgetting operation unit, the target forgetting data set is subjected to a forgetting operation through gradient ascent, including: assumed represents a model that has only been fine-tuned without forget training, Represents a model that has only been trained with gradient ascent and forgetfulness. For any sample in the target forgetting set , the parameter update of gradient ascent is expressed as: ； in, for The model parameters, is the learning rate, is the objective function of gradient ascent, In the model The gradient of For the updated Parameters, Indicates that the initial parameters Next, given input Output The conditional probability of .

7. The auxiliary system for enhancing large model forgetting by parameter extrapolation according to claim 5 or 6, characterized in that: In the parameter extrapolation unit, specifically, it includes: according to the optimal balanced model parameters and the parameter update vector , amplifying the parameter update vector by parameter extrapolation, including: ,in, Used to control the magnification of the parameter update vector, is the parameter update vector after parameter extrapolation and amplification.

8. A computer device, characterized in that: include: a processor and a computer readable storage medium; a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in any one of claims 1 to 4 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which is suitable for being loaded by a processor and executing the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in any one of claims 1 to 4.

10. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the auxiliary method for enhancing large model forgetting by parameter extrapolation as described in any one of claims 1 to 4.

Images