An interface interaction method, device and electronic equipment of a large model

Abstract

This application provides a method, apparatus, and electronic device for user interface interaction of large models, relating to the field of large models. It provides a location entry point for each input content within the input information in the contribution display interface. Thus, when a user identifies the input content that needs debugging through the contribution display interface, they can locate that content within the input information using the location entry point, resulting in high interaction efficiency and a good user experience. The method includes: displaying a first interface. The first interface includes a first window and a second window. The first window displays the input information of the large model, and the second window displays the input information and the contribution of each part of the input information, with the contribution indicating the degree of influence on the output result of the large model. In response to a selection operation on the first content in the second window, the first content is marked in the first window. The first content can be any content in the input information.

Description

Technical Field

[0001] This application relates to the field of large models, and more particularly to a method, apparatus, and electronic device for interface interaction of large models. Background Technology

[0002] A large language model (LLM), also known as a large language model, is a model that generates expected content based on a given prompt. The expected content can be the content corresponding to the prompt. For example, given prompts including examples of translating English into French, and the English text to be translated, the large model can generate the corresponding French text.

[0003] Large models are highly sensitive to cue words; changes to individual tokens within a cue word can have a significant impact on the model's output. In the realm of large models, contribution values ​​are typically used to quantify this impact. For example, if the overall contribution of the cue word can be defined as 1, then the contribution of each element within the cue word can be represented by a number less than 1.

[0004] To ensure that the output of a large model meets expectations, users typically need to debug the model's prompts based on contribution. However, this debugging process becomes quite cumbersome when the prompts contain a lot of information. Summary of the Invention

[0005] This application provides a method, apparatus, and electronic device for large-scale model interface interaction, which provides location entry points for each input content in the input information within the contribution display interface. Thus, when a user identifies the input content that needs debugging through the contribution display interface, they can locate that input content within the input information using the location entry points, resulting in high interaction efficiency and a better user experience.

[0006] Firstly, a method for user interface interaction of a large model is provided, comprising: displaying a first interface. The first interface includes a first window and a second window. The first window is used to display the input information of the large model, and the second window is used to display the input information and the contribution of each part of the input information, wherein the contribution indicates the degree of influence on the output result of the large model. In response to a selection operation on the first content in the second window, the first content is marked in the first window. The first content is any content in the input information.

[0007] Based on this scheme, when a selection operation is detected on the first item in the contribution display window, the first item in the input information can be marked in the input window, eliminating the process of the user searching in the input information, thereby improving interaction efficiency and enhancing the user experience.

[0008] Based on the interface interaction method of the large model provided in the first aspect, in some possible implementations, the first interface includes a first control. The first control is used to control the display method of the second window. The display method includes at least one of the following: displaying the contribution of each tag in the input information, displaying the contribution of each field in the input information, and displaying the contribution of each sentence group in the input information. Here, a sentence group includes multiple fields, and each field includes one or more tags. Based on this scheme, users can easily view the contribution of each part of the input information at different granularities.

[0009] Combining the interface interaction method of the large model provided in the first aspect, in some possible implementations, the second window displays the contribution of each first unit in the input information. The first unit is a label or a field composed of labels. The method also includes: in response to the selection of multiple first units in the second window, displaying the sum of the contributions of the multiple first units. Based on this scheme, users can easily view the overall contribution of any content, resulting in a better user experience.

[0010] Combining the interface interaction method of the large model provided in the first aspect, in some possible implementations, the second window displays the contribution of each second unit in the input information. The second unit is a field composed of tags or a sentence group composed of fields. The method also includes: in response to a selection operation on the second unit in the second window, displaying the contribution of each first unit in the second unit. Wherein, when the second unit is a field, the first unit is a tag. When the second unit is a sentence group, the first unit is either a tag or a field. Based on this scheme, users can easily view the overall contribution of any content, resulting in a better user experience.

[0011] Combining the interface interaction methods of the large model provided in the first aspect, some possible implementations include marking the first content in the first window, including: marking the first content in the first window with a first identifier. The first identifier includes any of the following: box, speech bubble, icon, cursor, color, and arrow. Based on this scheme, the first content that the user needs to find in the input information can be clearly displayed, omitting the process of the user searching in the input information, thereby improving interaction efficiency and enhancing the user experience.

[0012] Combining the interface interaction method of the large model provided in the first aspect, in some possible implementations, the first identifier is used as an input cursor to indicate the input position. Based on this scheme, it is convenient for users to modify the input information that needs to be modified, and the interaction efficiency is relatively high.

[0013] Secondly, a large-scale interface interaction device is provided, comprising multiple interacting modules for implementing any of the methods in the first aspect.

[0014] Thirdly, a large-scale model interface interaction device is provided, comprising a first module and a second module. The first module displays a first interface, which includes a first window and a second window. The first window displays the input information of the large model, and the second window displays the input information and the contribution of each part of the input information, whereby the contribution indicates the degree of influence on the output result of the large model. The second module, in response to a selection operation on the first content in the second window, marks the first content in the first window. The first content can be any content from the input information.

[0015] Fourthly, an electronic device is provided, comprising one or more processors. The one or more processors are configured to execute computer programs or instructions to implement the method of any of the first aspects.

[0016] Fifthly, an electronic device is provided, characterized in that it includes a memory and one or more processors. The memory is used to store computer programs or instructions. The one or more processors are used to execute the computer programs or instructions in the memory, causing the electronic device to perform the methods as described in any of the first aspects.

[0017] In a sixth aspect, a computer-readable storage medium is provided, comprising a computer program or instructions that, when executed, cause the method of any one of the first aspects to be implemented.

[0018] In a seventh aspect, a computer program product is provided, comprising a computer program or instructions that, when executed, cause the method of any one of the first aspects to be implemented.

[0019] Eighthly, a chip device is provided, including a processor and a memory. The processor is used to invoke a computer program or computer instructions stored in the memory to cause the processor to execute any of the implementations described in the first aspect. Optionally, the processor is coupled to the memory via an interface.

[0020] It should be understood that the second to eighth aspects of this application are consistent with or correspond to the technical solutions of the first aspect of this application, and the beneficial effects achieved by each aspect and the corresponding feasible implementation are similar, and will not be repeated here. Attached Figure Description

[0021] Figure 1 A schematic diagram of a debugging prompt word provided in an embodiment of this application;

[0022] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application;

[0023] Figure 3A flowchart illustrating a large-scale model interface interaction method provided in an embodiment of this application;

[0024] Figure 4 A schematic diagram of a user interface for a large model provided in an embodiment of this application;

[0025] Figure 5 A schematic diagram of a switching window provided in an embodiment of this application;

[0026] Figure 6 A schematic diagram illustrating the display of contribution level provided in an embodiment of this application;

[0027] Figure 7 A schematic diagram of a user interface for another large model provided in an embodiment of this application;

[0028] Figure 8 A schematic diagram of a contribution display window provided in an embodiment of this application;

[0029] Figure 9 A schematic diagram of an edit icon provided in an embodiment of this application;

[0030] Figure 10 A schematic diagram of a user interface for another large model provided in an embodiment of this application;

[0031] Figure 11 This is a schematic diagram of the structure of a large-scale interface interaction device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0033] References to "one embodiment" or "some embodiments" as described in this application mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0034] Furthermore, in the embodiments of this application, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as an "example" in this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the term "example" is intended to present concepts in a concrete manner. In the embodiments of this application, "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably, and it should be noted that their intended meanings are consistent unless their distinction is emphasized.

[0035] The application scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0036] To facilitate understanding, the background of the proposed embodiments of this application will be introduced below.

[0037] During the use of large models, it is often necessary to debug the prompts for inputting the large model. For ease of explanation, the prompts for inputting the large model will be referred to as input information in the following examples.

[0038] For example, please refer to Figure 1 This is a schematic diagram of a debugging prompt word provided in an embodiment of this application. Figure 1 As shown, the large model can receive input information and output results based on that input information. The results information can include the large model's output on the input information, as well as contribution information. The contribution information includes indicators used to indicate the content of each part of the input information (…). Figure 1 The contribution of the input content to the degree of influence on the output result.

[0039] For example, the overall contribution of the input information can be defined as 1, then the contribution of each input element can be represented by a number less than 1. In this embodiment, the contribution can be calculated using an attribution algorithm. For example, the attribution algorithm may include, but is not limited to, gradient-based attribution algorithms (such as Smooth Gradients algorithm, Integrated Gradients algorithm), occlusion-based attribution algorithms (such as Shapley value algorithm, Occlusion-patch algorithm), and layer-by-layer backpropagation-based attribution algorithms (such as Deep Taylor algorithm, DeepShap algorithm), etc. This embodiment does not specifically limit the method for generating the contribution.

[0040] In some scenarios, there may be a discrepancy between the output of the large model and the user's expectations. Therefore, users can adjust the prompts based on the contribution of each input element to narrow this gap. For example, a user can first identify the input elements with higher contributions, find them in the input information, and adjust them to achieve a coarse adjustment of the large model's output. Then, the user can identify the input elements with lower contributions, find them in the input information, and adjust them to achieve a fine adjustment of the large model's output.

[0041] However, when the input information includes a large amount of content, users may find it difficult to quickly and accurately locate the input content that needs to be debugged, making the debugging process complex and cumbersome.

[0042] To address the aforementioned issues, embodiments of this application provide a large-scale interface interaction method, apparatus, and electronic device, which provides location entry points for each input content within the input information in the contribution display interface. Thus, when a user identifies the input content requiring adjustment through the contribution display interface, they can locate that content within the input information using the location entry points, resulting in high interaction efficiency and a better user experience.

[0043] The interface interaction method for large models provided in this application can be applied to electronic devices that include large models. The electronic device may include a display capable of showing a user interface, such as a screen.

[0044] For example, electronic devices can be mobile phones, tablets, laptops, handheld computers, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, etc.

[0045] Please refer to Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 2 As shown, the electronic device 200 may include one or more processors 201. Figure 2 Taking a processor as an example) and one or more displays 202 coupled to the processor. Figure 2 (Taking a monitor as an example). The monitor 202 can implement the display functions involved in the interface interaction method of the large model provided in the embodiments of this application.

[0046] Optionally, the electronic device 200 may further include one or more memories 203 coupled to the processor 201. Figure 2 Taking a memory as an example (represented by a dashed box), the memory 203 is used to store computer programs or instructions and / or data, and the processor 201 is used to execute the computer programs or instructions and / or data stored in the memory 203, so that the methods or steps in the embodiments of this application are executed.

[0047] Alternatively, the memory 203 may be integrated with the processor 201 or set separately.

[0048] It should be understood that this is merely an exemplary description of the electronic device in the embodiments of this application. In some possible implementations, the electronic device may include, for example... Figure 2 The number of more or fewer devices shown is not limited here.

[0049] The following is a detailed description of the interface interaction method for large models provided in the embodiments of this application. It should be understood that the subject executing this method can be an electronic device. The electronic device can be a device or apparatus with a display and a chip, or a device or apparatus with a display and integrated circuits, etc., and is not limited thereto.

[0050] Please refer to Figure 3 This is a flowchart illustrating a large-model interface interaction method provided in an embodiment of this application. It should be understood that the following... Figure 3 The interface interaction methods of the large model shown are merely exemplary and do not represent that this application is limited thereto.

[0051] like Figure 3 As shown, the method may include the following steps.

[0052] S301, Display the first interface.

[0053] The first interface can refer to the user interface of the large model. This application does not limit the platform of the user interface; for example, the user interface can be a webpage, an interface in an integrated development environment (IDE), or an interface on another independent platform.

[0054] For example, please refer to Figure 4 This is a schematic diagram of a user interface for a large model provided in an embodiment of this application. Figure 4 As shown, the user interface 400 of this large model may include an input window 401 and a contribution display window 402.

[0055] Input window 401, also known as the first window, is used to display input information. This input information can consist of tokens, which can be words, letters, characters, etc. A token is the smallest unit in the input information used to indicate semantic or syntactic structure. For example, the word "eating" in the input information can be composed of the tokens "eat" and "ing".

[0056] In this embodiment, words or characters separated by spaces in the input information can be referred to as words. It should be understood that a word can consist of one or more tags. Furthermore, multiple words can be referred to as a component. For example, one or more lines of content in the input information can be referred to as a component.

[0057] The contribution display window 402, also known as the second window, is used to display contribution information. For an explanation of how to generate contribution information, please refer to the foregoing embodiments; it will not be repeated here. The contribution display window 402 can display the contribution of each tag, the contribution of each field, or the contribution of each sentence group; this application does not limit this to any particular type.

[0058] Optionally, the user interface may also include a toggle control. For example, such as Figure 4 As shown, the toggle control 403 ( Figure 4 (The area indicated by the dashed line) can be located in the contribution display window 402.

[0059] The toggle control can be used to switch the display method of contribution information. The display methods include, but are not limited to: displaying the contribution of each tag in the input information, displaying the contribution of each field in the input information, and displaying the contribution of each sentence group in the input information.

[0060] In some possible implementations, the toggle control may include options for switching between different display modes. For example, the toggle options may include three options: token, word, and component. The token option displays the contribution of each token in the input information; the word option displays the contribution of each field in the input information; and the component option displays the contribution of each sentence group in the input information.

[0061] The electronic device implementing this method can, in response to a user's selection of a switching option, switch the display mode of the contribution information in the contribution display window 403 to the display mode corresponding to that switching option. The selection operation can be a click, double-click, long press, etc., and is not limited here.

[0062] In some other possible implementations, the electronic device can respond to user requests. Figure 4 The selection operation of the switch control 403 is displayed as follows: Figure 5 The switching window 501 shown can include options for switching between different display modes. Figure 5 (Taking the three options of token, word, and component as an example).

[0063] It should be understood that when switching options include token, word, and component, the electronic device can respond to the user's selection of the token option (such as a click) and display the contribution of each token in the input information. The electronic device can also respond to the user's selection of the word option and display the contribution of each field in the input information. Furthermore, the electronic device can respond to the user's selection of the component option and display the contribution of each sentence group in the input information.

[0064] As can be seen, the first interface provided in this application embodiment supports multiple display methods for contribution information. Among these multiple display methods, contribution display is based on prompt word structure from high-level to low-level, which facilitates users to efficiently view contribution information and thus enables them to quickly locate content with large or small contributions by moving from the overall picture to the details.

[0065] Additionally, in some possible implementations, the contribution information displayed in the contribution display window may include the individual contributions of multiple first units (such as tags, fields, sentence groups, etc.). The electronic device can respond to a selection operation on these multiple first units by displaying the overall contribution of those first units (i.e., the sum of their contributions).

[0066] For example, such as Figure 6 As shown, the contribution information displayed in the contribution display window includes a contribution of 0.2 for input content a, a contribution of 0.1 for input content b, and a contribution of 0.05 for input content c. Therefore, the electronic device can respond to selection operations for input content a, input content b, and input content c. Figure 6 Taking the selection operation as an example, the contribution score is displayed as 0.35 (that is, the sum of the contributions of input content a, input content b, and input content c). Input content a, input content b, and input content c can be consecutive or non-consecutive; no restriction is placed here.

[0067] Optionally, when displaying the overall contribution of multiple contents, the electronic device can, in response to a return operation, or after a preset time, or in response to a selection operation of the multiple contents, resume displaying the individual contribution of each content. The preset time can be set as needed, such as 3 seconds, 5 seconds, etc. Furthermore, the return operation can be a click operation on a blank area, a selection operation on a return control in the first interface, etc. This application embodiment does not limit these aspects.

[0068] In some possible implementations, the user interface 400 of the large model may also include an output window 404. This output window 404 can be used to display the output results of the large model in response to the input information.

[0069] In some possible implementations, the user interface of the large model may also include a launch control. Figure 4 (Not shown in the image). This startup control is used to instruct the large model to generate and display the output results based on the input information.

[0070] In some possible implementations, the user interface of the large model may also include a parameter setting window. Figure 4 (Not shown in the image). The parameter setting window may include one or more parameter setting options or controls for the large model, such as a control to display the contribution display window, an option to limit the size of the input information, and a temperature setting option for the large model. For example, the parameter setting window may include a contribution switch control. The contribution switch control may include an on state and an off state. When the contribution switch is on, the contribution display window may be displayed in the first interface. When the contribution switch is off, the contribution display window may not be displayed in the first interface. It should be understood that this is merely an illustrative example and does not represent a limitation of this application.

[0071] S302, In response to the selection operation of the first content in the contribution display window, the first content in the input information is marked in the input window.

[0072] The selection operation can be a single click, double click, or swipe selection. The first content refers to any content within the contribution information, such as one or more tags, one or more fields, or one or more sentence groups. This application does not limit this.

[0073] Furthermore, this application's embodiments include multiple selection operations. The selection operation for a specific content includes all possible operations capable of selecting that content. In this application's embodiments, each selection operation can refer to the same operation or different operations; no limitation is made here. For example, Figure 6The selection operation shown can be a box selection operation, and the selection operation in S302 can be a click operation.

[0074] In some possible implementations, indicating the first content can mean highlighting the first content. For example, the first content can be displayed using a special color (black can be used to display the content of the input information, while blue, red, or other colors can be used to highlight the first content). Another example is highlighting the first content using boxes, arrows, speech bubbles, or similar methods.

[0075] In some other possible implementations, indicating the first content can also refer to moving the edit icon to the location of the first content. Here, the edit icon can also be called the edit cursor or input cursor, used to indicate the input position. In other words, the edit icon can indicate the position where the input content is displayed.

[0076] It should be understood that after viewing the contribution information in the contribution display window, users may adjust the input information based on the contribution of each part of the content. In the interface interaction method of the large model provided in this application embodiment, when the electronic device detects a selection operation on the first content in the contribution display window, it can mark the first content in the input information in the input window, omitting the process of the user searching in the input information, thereby improving interaction efficiency and enhancing the user experience.

[0077] To more clearly illustrate the interface interaction method for large models provided in the embodiments of this application, two specific applications of the method are described below with reference to the accompanying drawings.

[0078] Please refer to Figure 7 This is a schematic diagram of a user interface for another large model provided in an embodiment of this application. Figure 7 As shown, the user interface 700 includes an input window 701, an output window 702, and a contribution display window 703. The input information for the large model is the content in the input window 701. The output information for the large model is the content in the output window 702. The contribution information is the content in the contribution display window 703 (blank boxes represent line breaks).

[0079] Figure 7In the input information, the large model is instructed to translate English into French. The input information includes "##Persona:", "##Instruction:", and "##Examples:", which indicate the structure of the input information (i.e., prompts). Specifically, the line containing "##Persona:" indicates the role of the large model, the line containing "##Instruction:" indicates the instruction the large model executes, and the line containing "##Examples:" indicates examples. "English:I don'tlike eating tomato^" indicates the English to be translated. The "^" marker indicates that the large model should output the corresponding French translation of the English.

[0080] Optionally, the contribution display window 703 may also include a toggle control 704. For a description of the toggle control, please refer to the description in the foregoing embodiments; it will not be repeated here. Figure 7 As shown, the switching control indicates that the current display mode is word, that is, the contribution display interface displays the contribution of each field in the input information.

[0081] In some possible implementations, the electronic device may display, in response to a selection operation on the toggle control 704, a display such as Figure 5 The switching window 501 is shown. Then, in response to the selection of the component option, the electronic device can modify the display method of the contribution information in the contribution display window 703 to be as follows: Figure 8 As shown, this displays the contribution of each sentence group in the input information. Specifically, "##Persona_0" corresponds to "##Persona:You are a translation expert.", "##Instruction_0" corresponds to "##Instruction:Please translate the sentence to French:", "##Examples_0" corresponds to "##Examples:English:I love you^French:Je t'aime", and "##Examples_1" corresponds to "##Examples:English:I am happy^French:Je suis content" and "English:I don't like eating tomato^".

[0082] It should be noted that the division of sentence groups here is merely exemplary and does not imply that this application is limited to this.

[0083] In some possible scenarios, users may need to adjust input information (such as modifying the given example). Electronic devices can also respond to... Figure 7 in "##Examples:English:I love you^French:Je t'aime" or Figure 8 The selection operation of "##Examples_0" is performed in... Figure 7 The input window 701 shown is labeled "##Examples:English:I love you^French:Je t'aime". For example, as... Figure 9 As shown, the edit icon 901 is displayed in the input window 701 at the location of "##Examples:English:I love you^French:Je t'aime".

[0084] Please refer to Figure 10 This is a schematic diagram of a user interface for another large model provided in an embodiment of this application. Figure 10 As shown, the user interface 1000 includes an input window 1001, an output window 1002, and a contribution display window 1003. The input information for the large model is the content in the input window 1001. The output information for the large model is the content in the output window 1002. The contribution information is the content in the contribution display window 1003 (0x0A represents a newline character).

[0085] Figure 10 In this example, the input information instructs the large model to label the book as positive or negative based on the input content. It can be seen that the large model labeled "The book was very touching and the plot was gripping:" as negative (it should actually be positive), which is not as expected. In this case, the user can view the contribution information from highest to lowest. It can be found that "IADMIRED the book" in the example should correspond to a positive label, while the original input was a negative label, which needs to be modified. The user can modify this by selecting "I ADMIRED the book: negative" and marking it in input window 1001 (changing "negative" to "positive"). Thus, the large model's output is positive, which is as expected.

[0086] in addition, Figure 10 The user interface 1000 shown may also include a toggle control. Figure 10(Not shown). For a description of the switching control, please refer to the description in the foregoing embodiments, which will not be repeated here.

[0087] This application also provides a large-scale model interface interaction device, which may include multiple interacting modules for implementing any of the methods described in the above method embodiments.

[0088] For example, please refer to Figure 11 This is a structural schematic diagram of a large-scale interface interaction device provided in an embodiment of this application. Figure 11 As shown, the interface interaction device 1100 of the large model includes: a first module 1101 and a second module 1102.

[0089] The first module 1101 is used to display a first interface. The first interface includes a first window and a second window. The first window is used to display the input information of the large model, and the second window is used to display the input information and the contribution of each part of the input information. The contribution indicates the degree of influence on the output result of the large model. The second module 1102 is used to mark the first content in the first window in response to the selection operation of the first content in the second window. The first content is any content in the input information.

[0090] In some possible implementations, the first interface includes a first control. The first control controls the display method of the second window. The display method includes at least one of the following: displaying the contribution of each tag in the input information, displaying the contribution of each field in the input information, and displaying the contribution of each sentence group in the input information. Wherein, a sentence group includes multiple fields, and each field includes one or more tags.

[0091] In some possible implementations, the second window displays the contribution of each first unit in the input information. The first unit is a tag or a field composed of tags. The second module 1102 is also configured to display the sum of the contributions of the multiple first units in response to a selection operation on multiple first units in the second window.

[0092] In some possible implementations, the second window displays the contribution of each second unit in the input information. The second unit is either a field composed of tags or a group of sentences composed of fields. The second module 1102 is also configured to display the contribution of each first unit within the second unit in response to a selection operation on a second unit in the second window. Wherein, when the second unit is a field, the first unit is a tag. When the second unit is a group of sentences, the first unit is either a tag or a field.

[0093] In some possible implementations, the second module 1102 is also used to mark the first content in the first window by a first identifier. The first identifier includes any of the following: a box, a bubble, an icon, a cursor, a color, or an arrow.

[0094] This application also provides an electronic device including one or more processors. The one or more processors are used to execute computer programs or instructions to implement any of the methods described in the above method embodiments.

[0095] This application also provides a computer-readable storage medium, which includes a computer program or instructions that, when executed, enable the implementation of any of the methods described in the above embodiments.

[0096] This application also provides a computer program product, which includes a computer program or instructions, such that when the computer program or instructions are run, the method of any one of the above method embodiments is implemented.

[0097] This application also provides a chip device including a processor and a memory. The processor is used to invoke a computer program or computer instructions stored in the memory to cause the processor to execute any implementation of the above-described method embodiments. Optionally, the processor is coupled to the memory via an interface.

[0098] It should be noted that some optional features in the various embodiments of this application may not depend on other features in certain scenarios, or may be combined with other features in certain scenarios, without limitation.

[0099] The solutions in the various embodiments of this application can be used in reasonable combinations, and the explanations or descriptions of various terms, similar operations, or steps appearing in the embodiments can be referenced or explained to each other in the various embodiments, without limitation.

[0100] This application also provides a system, including one or more of the above-described devices, apparatuses, computer-readable storage media, computer program products, chips, or chip systems.

[0101] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the explanation of the relevant content and beneficial effects of any of the devices, equipment, and media provided above can be referred to the corresponding method embodiments provided above, and will not be repeated here.

[0102] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0103] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0104] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0105] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the part of the technical solution that makes an essential contribution, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0106] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for user interface interaction in a large model, characterized in that, include: The first interface is displayed; the first interface includes a first window and a second window; The first window is used to display the input information of the large model, and the second window is used to display the input information and the contribution of each part of the input information, wherein the contribution is used to indicate the degree of influence on the output result of the large model; In response to a selection operation on the first content in the second window, the first content is marked in the first window; the first content is any content in the input information.

2. The method according to claim 1, characterized in that, The first interface includes a first control; the first control is used to control the display mode of the second window; the display mode includes at least one of the following: displaying the contribution of each tag in the input information, displaying the contribution of each field in the input information, and displaying the contribution of each sentence group in the input information; wherein, the sentence group includes multiple fields, and the fields include one or more tags.

3. The method according to claim 1, characterized in that, The second window displays the contribution of each first unit in the input information; The first unit is a tag or a field composed of said tags; the method further includes: In response to a selection operation on multiple first units in the second window, the sum of the contributions of the multiple first units is displayed.

4. The method according to claim 1, characterized in that, The second window displays the contribution of each second unit in the input information; the second unit is a field composed of tags or a group of sentences composed of the fields; The method further includes: In response to a selection operation on a second unit in the second window, the contribution of each first unit in the second unit is displayed; wherein, when the second unit is the field, the first unit is the tag; when the second unit is the sentence group, the first unit is the tag or the field.

5. The method according to any one of claims 1-4, characterized in that, The step of marking the first content in the first window includes: The first content in the first window is marked by a first identifier; the first identifier includes any one of the following: box, bubble, icon, cursor, color, arrow.

6. The method according to claim 5, characterized in that, The first identifier is an input cursor used to indicate the input position.

7. A large-scale model interface interaction device, characterized in that, It includes multiple interacting modules for implementing the method as described in any one of claims 1 to 6.

8. An electronic device, characterized in that, It includes one or more processors; the one or more processors are configured to execute computer programs or instructions to implement the method of any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a computer program or instructions that, when executed, cause the method of any one of claims 1-6 to be implemented.

10. A computer program product, characterized in that, The computer program product includes a computer program or instructions that, when executed, cause the method described in any one of claims 1-6 to be implemented.

Images