Information processing method and device, electronic equipment and storage medium

Abstract

The present disclosure provides an information processing method, relates to the technical field of artificial intelligence, and particularly relates to the fields of large models, agents and interaction technologies. A specific implementation scheme is as follows: in the process of interacting with a large model, input information in the process of performing thinking processing based on initial question information is detected; in response to detecting the input information, second intermediate thinking content is determined according to the first intermediate thinking content and the input information, the first intermediate thinking content being thinking content that has been generated in the process of performing thinking processing based on the initial question information; and response information for at least one of the initial question information and the input information is generated according to the second intermediate thinking content. The present disclosure also provides an information processing device, an electronic equipment and a storage medium.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more particularly to the fields of large models, intelligent agents, and interaction technologies. More specifically, this disclosure provides an information processing method, apparatus, electronic device, and storage medium. Background Technology

[0002] With the development of artificial intelligence technology, the application of intelligent agents and large models is constantly increasing. The thinking process of large models or intelligent agents can be continuously carried out according to preset algorithms and processes. Summary of the Invention

[0003] This disclosure provides an information processing method, apparatus, device, and storage medium.

[0004] According to one aspect of this disclosure, an information processing method is provided, the method comprising: detecting input information during the process of performing thinking processing based on initial question information in the process of interacting with a large model; in response to detecting the input information, determining a second intermediate thinking content based on a first intermediate thinking content and the input information, the first intermediate thinking content being thinking content generated during the process of performing thinking processing based on the initial question information; and generating response information for at least one of the initial question information and the input information based on the second intermediate thinking content.

[0005] According to another aspect of this disclosure, an information processing apparatus is provided, the apparatus comprising a detection module for detecting input information during the process of performing thinking processing based on initial question information during interaction with a large model; a determination module for determining second intermediate thinking content based on first intermediate thinking content and input information, the first intermediate thinking content being thinking content generated during the process of performing thinking processing based on initial question information; and a generation module for generating response information for at least one of the initial question information and input information based on the second intermediate thinking content.

[0006] According to another aspect of this disclosure, an intelligent agent is provided, comprising: a receiving module for receiving information to be processed; a processing module for obtaining output information by calling a large model to execute an information processing method as provided in this disclosure based on the information to be processed received by the receiving module; and an output module for outputting the output information obtained by the processing module.

[0007] According to another aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a method provided according to this disclosure.

[0008] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing a computer to perform the methods provided according to this disclosure.

[0009] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to this disclosure.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0011] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0012] Figure 1 This is an exemplary system architecture diagram of an embodiment of the present disclosure in which information processing methods and apparatus can be applied;

[0013] Figure 2 This is a flowchart of an information processing method according to an embodiment of the present disclosure;

[0014] Figure 3A This is a schematic diagram of a pre-pause visual interface according to an embodiment of the present disclosure;

[0015] Figure 3B This is a schematic diagram of the visual interface after pause in one embodiment of this disclosure;

[0016] Figure 4A This is a schematic diagram of a visual interface for receiving input question information according to an embodiment of the present disclosure;

[0017] Figure 4B This is a schematic diagram of inserting an input question node according to an embodiment of the present disclosure;

[0018] Figure 4C This is a schematic diagram illustrating the adjusted thinking result obtained according to an embodiment of the present disclosure;

[0019] Figures 5A to 5D This is a schematic diagram of another visual interface according to this disclosure;

[0020] Figure 6 This is a block diagram of an information processing apparatus according to an embodiment of the present disclosure;

[0021] Figure 7 This is a block diagram of an intelligent agent according to an embodiment of the present disclosure; and

[0022] Figure 8 This is a block diagram of an electronic device to which an information processing method can be applied, according to an embodiment of the present disclosure. Detailed Implementation

[0023] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0024] As mentioned above, when performing tasks, the agent's thought process can continuously run according to preset algorithms and procedures, lacking flexible and dynamic interaction mechanisms. Users find it difficult to intervene during the thought process. When new needs arise or deviations are discovered in the thought process, the agent cannot adjust its thinking process in a timely manner, resulting in output responses that may not meet expectations and making it difficult to adapt to complex and ever-changing application scenarios.

[0025] In some intelligent agents or large models, the thinking and execution processes are often conducted in a "one-time continuous run" mode, lacking a reliable pausing mechanism. The thinking process cannot be flexibly paused, and the user can only passively wait for the results. When the agent's thinking direction deviates from the user's expectations (such as generating irrelevant content or analyzing erroneous data), the user cannot pause midway or interrupt redundant calculations; they can only wait for the agent to complete the entire process before restarting, resulting in wasted computing resources and time. Some agents support forced termination, but this leads to the loss of intermediate thinking processes, context information, and computational state. If the user wants to adjust the task, they must start from scratch and cannot continue optimizing or calculating based on the existing progress.

[0026] Furthermore, during the agent's thought process, users cannot supplement their needs in real time, resulting in significant interaction lag. User input is mostly limited to the initial stage of the task, lacking a real-time interactive interface during operation. If a user generates new ideas (such as adding details, adjusting goals, or asking follow-up questions) while the agent is running, these new ideas and needs cannot be promptly and directly transmitted to the agent. They can only be initiated again after the current task is completed, leading to a disconnect between the current task's outcome and the user's dynamic needs. Taking the agent's analysis of market data as an example, during the analysis process, the user may suddenly want to add a new analytical dimension, but cannot inform the agent of this new dimension in real time, and must resubmit the complete task. Even if some agents support subsequent input, it is difficult to associate new requirements with previous thought processes. The agent may ignore historical context, leading to repetitive work or logical contradictions.

[0027] Furthermore, the agent's thought processes are difficult to edit, and error correction is costly. If the agent's intermediate reasoning contains errors, the user can point out the problem through the final result, but cannot directly modify the erroneous steps or logic in the thought process. The user can also re-enter instructions to make the agent deduce from scratch, leading to repetitive work. For example, when solving equations, if a formula is called incorrectly in a certain step, the user cannot directly correct that step; they can only input instructions to inform the agent of the calculation error so that the agent can recalculate. During the recalculation process, the agent may repeat the same error.

[0028] In the absence of an effective pause mechanism, the inability to supplement requirements in real time, and the difficulty in editing thought processes, these agents perform poorly in complex and dynamic scenarios. For tasks requiring repeated adjustments, users must start and stop the agent multiple times, regenerating the complete context each time. This process is cumbersome, error-prone, and involves multiple rounds of tasks, resulting in low efficiency. Users are constantly in a passive loop of "waiting for results – discovering problems – re-entering," unable to deeply participate in the agent's thought process, making it difficult to establish a sense of "collaboration," and leading to a fragmented user experience. The accuracy of the final result highly depends on the completeness of the initial instructions, placing extremely high demands on the user's expressive abilities.

[0029] Therefore, in order to effectively improve thinking efficiency and enhance user experience, this disclosure provides an information processing method, which will be described below.

[0030] Figure 1 This is a schematic diagram of an exemplary system architecture to which information processing methods and apparatus can be applied according to an embodiment of this disclosure. It should be noted that... Figure 1 The examples shown are merely examples of system architectures that can be applied to the embodiments of this disclosure, in order to help those skilled in the art understand the technical content of this disclosure, but do not mean that the embodiments of this disclosure cannot be used in other devices, systems, environments or scenarios.

[0031] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a terminal device 102, a network 103, and a server 104. The network 103 serves as a medium for providing a communication link between the terminal device 102 and the server 104. The network 103 may include various connection types, such as wired and / or wireless communication links, etc.

[0032] Users can interact with server 104 via network 103 using terminal device 102. Users can send problem information 101 through the interactive interface provided by terminal device 102. Terminal device 102 can then send problem information 101 to server 104 via network 103, causing server 104 to invoke an intelligent agent to process problem information 101 and output a response result 105. The server then sends the response information 105 back to terminal device 102, allowing terminal device 102 to display the response information 105 to the user.

[0033] The terminal device 102 can be equipped with various communication client applications, such as smart assistant applications, knowledge reading applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software (for example only). Users can input question information 101 into the interactive interface of these client applications, and these client applications will display the generated response information 105 to the user.

[0034] Terminal device 102 can be configured with various electronic devices that have a display screen and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0035] Server 104 can be a server providing various services, such as a backend management server supporting the content browsed by the user through the interactive interface of terminal device 102 (for example only). The backend management server can invoke the intelligent agent to query data upon receiving the problem information, and feed back the response results to terminal device 102, which will then display them through the interactive interface. For example, server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud computing, network services, and middleware services.

[0036] It should be noted that the information processing method provided in this embodiment can generally be executed by server 104. Correspondingly, the information processing device provided in this embodiment can also be located in server 104. The information processing method provided in this embodiment can also be executed by a server or server cluster that is different from server 104 and capable of communicating with terminal device 102 and / or server 104. Correspondingly, the information processing device provided in this embodiment can also be located in a server or server cluster that is different from server 104 and capable of communicating with terminal device 102 and / or server 104.

[0037] Alternatively, the information processing method provided in this embodiment can generally be executed by the terminal device 102. Accordingly, the information processing apparatus provided in this embodiment can generally be disposed in the terminal device 102.

[0038] As can be understood, the system architecture of this disclosure has been described above, and the method of this disclosure will be described below.

[0039] Figure 2 This is a flowchart of an information processing method according to an embodiment of the present disclosure.

[0040] like Figure 2 As shown, the method 200 may include operations S210 to S230.

[0041] When operating S210, during the interaction with the large model, the input information is detected during the thinking process based on the initial problem information.

[0042] In this embodiment of the disclosure, the large model can be a single-modal or multi-modal large model, capable of receiving problem information from various modalities. These multiple modalities may include text, language, audio, and video modalities.

[0043] In this embodiment of the disclosure, the question information can be information of any modality, and can include information of one or more modalities. For example, the question information can include one or more modalities of text information, image information, audio information, and video information. Taking text information as an example, the question information can be a question (query) provided by the user, such as "I am a junior high school physics teacher and I am preparing to teach the chapter on 'friction.' Please refer to the course outline in my cloud drive and help me organize the course ideas and make a teaching presentation."

[0044] In this embodiment of the disclosure, during interaction with the large model, the user can provide initial question information to the large model. After receiving the question information, the large model can perform thought processing based on the question information to generate thought content.

[0045] In this embodiment of the disclosure, the interaction process with the large model may include a single-round interaction. A single-round interaction may include: the large model receiving initial question information provided by the user and returning response information to the user in response to the initial question information. Before returning the response information in response to the initial question information, i.e., during the single-round interaction, it is detected whether input information has been received.

[0046] In operation S220, in response to the detection of input information, a second intermediate thought content is determined based on the first intermediate thought content and the input information.

[0047] In this embodiment of the disclosure, the first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information. For example, after receiving input information, the generated part of the thinking content can be used as the first intermediate thinking content. The first intermediate thinking content may include: "You are a junior high school physics teacher and are preparing a lesson on the chapter 'Friction.' You hope I can help you turn the lecture notes in the cloud drive into vivid courseware. To do this well, I will take it step by step, let's complete the lesson preparation together."

[0048] In this embodiment of the disclosure, the input information can be information of any modality, and may include information of one or more modalities. For example, the input information may include one or more modalities of text information, image information, audio information, and video information. As another example, the input information may be a follow-up question or editing information on the first intermediate thinking content. Taking a follow-up question as an example, the follow-up question may be "Please write another practice question for me." Based on this follow-up question, the aforementioned first intermediate thinking content can be adjusted to obtain second intermediate thinking content.

[0049] In operation S230, response information is generated based on the second intermediate thinking content, targeting at least one of the initial question information and the input information.

[0050] For example, based on the second intermediate thought process, further processing can be performed to generate response information for the initial question and follow-up questions. This response information could be, for example, a presentation file.

[0051] Through the embodiments of this disclosure, in the thinking process based on initial problem information, input information can be received, and new thinking content can be generated based on the input information and the already generated thinking content. This allows for timely responses to new demands and timely modification and correction of the thinking content, effectively reducing interaction lag, achieving flexible dynamic interaction, improving the interactive experience of intelligent agents and large models in various scenarios, and obtaining more accurate and demand-responsive results in a single dialogue round. This significantly improves problem-solving efficiency and reduces the consumption of computing resources.

[0052] As can be understood, the methods of this disclosure have been described above. In some embodiments, in order to effectively utilize hardware computing resources, the thinking process based on the initial problem information can be paused, as will be described below.

[0053] Figure 3A This is a schematic diagram of a pre-pause visual interface according to an embodiment of the present disclosure.

[0054] like Figure 3AAs shown, the visual interface i300 can display the initial question information q301 provided by the user. The initial question information q301 can be: "I am a junior high school physics teacher and I am preparing to teach the chapter on 'friction.' Please refer to the course outline in my cloud drive, help me organize the course ideas, create a teaching courseware, and make a demonstration animation that can intuitively show the motion process of the car on the inclined plane and display the relationship curve of velocity, displacement and time in real time, which is suitable for classroom demonstration."

[0055] The large model can process the initial problem information q301. This processing can include one or more thought steps. The visual interface includes information about the thought process. This information can include one or more thought nodes. A thought node can represent a thought step. For example... Figure 3A As shown, the multiple thinking steps for the initial problem information q301 can include "skill planning," "planning execution," and "thinking completion." The thinking process can be as follows: Figure 3A The first thought process information g301 shown represents this. The first thought process information g301 includes multiple thought nodes. These multiple thought nodes include skill planning nodes, planning execution nodes, and thought completion nodes. Through this embodiment of the disclosure, displaying the thought process information on a visual interface can intuitively demonstrate the thought process of a large model, making it easier for users to determine whether the large model's thinking is correct in key and important processing steps. This can effectively improve user experience and increase problem-solving efficiency.

[0056] In addition, the visual interface i300 can also display an input box ib30. The input box ib30 can receive input information provided by the user. The visual interface i300 also displays a pause control p301. When the pause control p301 is triggered, a pause command can be generated. It can be understood that the pause command can also be generated after the user enters pause information through the input box ib30. The pause information can be, for example, "Please pause thinking".

[0057] In some embodiments, the above method may further include: pausing the thought process based on the initial question information in response to receiving at least one of a pause command and input information via a visual interface. For example, during the skill planning step of the thought process, if the user believes that some information is missing in the initial question information q301, the pause control p301 can be triggered to generate a pause command. After receiving the pause command, the large model can pause the thought process based on the initial question information p301. The following will combine... Figure 3B Please provide an explanation.

[0058] Figure 3B This is a schematic diagram of a paused visual interface according to an embodiment of this disclosure.

[0059] like Figure 3BAs shown, when the pause control p301 is triggered, it switches to the continue control p302. The visual interface i300 can also display the terminate control e303. Triggering the terminate control e303 can end the thinking process based on the initial problem information q301. Triggering the continue control p302 can continue the thinking process.

[0060] In some embodiments, the above method may further include: saving the thought content generated during the thought processing based on the initial problem information as first intermediate thought content. For example... Figure 3B As shown, during the "Skill Planning" step in the thinking process, the first part of the thinking result t311 that has been generated can be: "You are a junior high school physics teacher, preparing a lesson on the chapter 'Friction,' and hope I can help you turn the lecture notes in your cloud drive into vivid courseware, and then create an animation demonstration that students will find interesting and easy to understand, showing a car running on an inclined plane, and showing the real-time changes in speed and displacement curves. To do this well, I will take it step by step, let's complete the lesson preparation together: Finding the lecture notes outline: I will carefully search for documents with filenames containing 'friction,' 'inclined plane motion,' or 'teaching notes' in the location of your specified cloud drive." Outline Analysis: After finding this document, I will organize and design the structure of the PPT to ensure that the resulting courseware fully matches your original teaching ideas and pace, without deviating from the intended path. Writing Teaching Materials: Considering the intuitive readability of your explanations of knowledge points in class, I will use professional PPT generation tools to design this PPT courseware for you under the guidance of your lecture outline. Creating Demonstration Animations: You mentioned demonstration animations in your request, presumably because you hope to allow students to concretely experience abstract knowledge in a more intuitive and engaging way. I will use professional code generation tools to design the scene and create the animation. Next, I will first go to your cloud drive to find your lecture outline. The first part of the thinking result t311 can be saved as the first intermediate thinking content. The first intermediate thinking content can be displayed on the visual interface. Through this disclosed embodiment, when the user generates new needs or needs to adjust the thinking content, the thinking can be paused in time, interrupting the redundant calculation of the large model, effectively improving the utilization rate of computing resources, effectively reducing time costs, and making the thinking direction of the large model more in line with the user's expectations. In addition, saving the generated thinking content can effectively preserve the intermediate thinking process, context information, and calculation status.

[0061] It is understood that the above description uses the example of pausing after the pause control is triggered to illustrate this disclosure. However, this disclosure is not limited to this; it can also pause thinking processing in response to receiving input information. Thus, in response to receiving at least one of a pause signal and input information, thinking processing based on initial question information is paused.

[0062] Understandably, the above text has explained some methods for handling pauses in thinking. The following text will explain the first intermediate thinking content of this disclosure.

[0063] In some embodiments, the first intermediate thinking content may include at least one of at least one intermediate thinking result and at least one partial thinking result. An intermediate thinking result may be the result obtained after completing a thinking step, while a partial thinking result may be the result obtained during the execution of the thinking step. For example... Figure 3B As shown, the first intermediate thinking content may include the first part of the thinking results t311 obtained during the execution of the skills planning steps.

[0064] In some embodiments, at least one of the intermediate thinking results and the partial thinking results includes editable material. For example, an intermediate thinking result may include one or more editable materials. A partial thinking result may also include one or more editable materials. Figure 3B As shown, editable material can be one or more characters that can be selected by the user. Alternatively, editable material can be any modality of the thought process, such as images or audio. Images can be directed acyclic graphs (DAGs). This DAG can, for example, indicate one or more sub-steps to be executed in the "planning execution" process.

[0065] As can be understood, the first intermediate considerations of this disclosure have been explained above, and the input information of this disclosure will be explained below.

[0066] In some embodiments, the input information may include at least one of the following information received during the performance of thought processing: input editing information for the target editable material; and input question information.

[0067] The input question information can be, for example, the follow-up question mentioned above, and can be received via an input module displayed on a visual interface. The input question information can indicate at least one thought result to be adjusted in the first intermediate thought content. For example, the input module can be the input box ib30 mentioned above. The thought result to be adjusted can be an intermediate thought result or a partial thought result.

[0068] The target editable material is the selected editable material among multiple editable materials. The input editing information can be one or more. Each input editing information corresponds to a target editable material. For example, the target editable material can be one or more characters selected by the user. Through the embodiments of this disclosure, the thinking direction of the intelligent agent or large model can be effectively adjusted in a round of dialogue based on various methods, which can effectively improve the user experience.

[0069] As you can understand, the input information for this disclosure has been explained above. The following will use the example of inputting question information to further illustrate this disclosure.

[0070] Figure 4A This is a schematic diagram of a visual interface for receiving input question information according to an embodiment of the present disclosure.

[0071] like Figure 4A As shown, the visual interface i400 can display initial problem information q401, first thinking process information g401, first part of thinking result t411, and input box ib40. It can be understood that the above descriptions of the visual interface i300, initial problem information q301, first thinking process information g301, first part of thinking result t311, and input box ib30 also apply to the initial problem information q401, first thinking process information g401, first part of thinking result t411, and input box ib40; these will not be repeated here.

[0072] During interaction with the large model, it's possible to detect whether input information has been received. After the user triggers the pause control, the user can click on input box ib40 to bring it up, allowing it to receive input. After bringing up input box ib40, a send control p303 can be displayed within it. The user can enter follow-up questions in input box ib40, such as "Please write another exercise for me." Next, the user can trigger send control p303 to provide the input question to the large model, allowing the input to be detected. Then, the above operation S220 can be executed.

[0073] In some embodiments of operation S220 described above, determining the second intermediate thought content may include: determining the intent recognition result of the input question information. For example, various tools and methods can be used to perform intent recognition on the input question information to obtain the intent recognition result.

[0074] In some embodiments, determining the intent recognition result of the input question information may further include: inserting an input question node into the thinking process information. The intent recognition result is displayed in the visible area of ​​the input question node in the visual interface. The input question node may represent a first additional step in the intent recognition of the input question information. For example, the aforementioned first intermediate thinking content may be displayed in the visible area of ​​the skill planning node. Before, during, or after executing the first additional step, some or all of the intent recognition results may be displayed in the visible area of ​​the input question node. The following will combine... Figure 4B Please provide an explanation.

[0075] Figure 4B This is a schematic diagram of inserting an input question node according to an embodiment of the present disclosure.

[0076] like Figure 4B As shown, after the agent receives the input question information, the visual interface i400 can display the input question information q402 and the second thinking process information g402. The second thinking process information g402 can be obtained by inserting the input question node into the second thinking process information g401. The input question node can be as follows: Figure 4B The question nodes shown.

[0077] After inserting the input question node, the first new step represented by the input question node can be executed to perform intent recognition on the input question information q402, obtaining the intent recognition result int401. The intent recognition result int401 can be "Based on your request 'Please write another exercise for me,' I will revise the action plan and add the step of generating exercise questions so that you can use it to assist students' classroom learning," which can be displayed in the visible area used for the input question node. Next, the above operation S220 can be continued.

[0078] In some embodiments, in some implementations of the above-described operation S220, determining the second intermediate thinking content may further include: adjusting at least one thinking result to be adjusted indicated by the input question information in the first intermediate thinking content according to the intent recognition result, to obtain the second intermediate thinking content. The second intermediate thinking content includes: the adjusted thinking result obtained by adjusting the thinking result to be adjusted. Through the embodiments of this disclosure, the generated thinking content can be adjusted in a targeted manner based on the intent of the follow-up information, making the thinking direction of the large model more in line with user expectations. In addition, based on the follow-up information, the complexity of adjusting the thinking content can be reduced, further improving the user experience.

[0079] For example, the intent recognition result int401 mentioned above includes the adjustment instruction information "replan the action plan". Therefore, the large model can use some of the thinking results generated during the execution of the above skill planning steps as thinking results to be adjusted.

[0080] Some thought processes can include at least one intermediate sub-result. The above "Finding the lecture outline: I will carefully search for documents with filenames containing 'friction,' 'inclined plane motion,' or 'teaching notes' in your specified cloud drive location" can be considered an intermediate sub-result. Similarly, the above "Outline analysis: After finding this document, I will organize and design the PPT structure to ensure the final presentation perfectly matches your original teaching approach and pace, staying on track" can also be considered an intermediate sub-result. It is understandable that intermediate thought processes can include at least one intermediate sub-result.

[0081] In some embodiments, adjusting at least one thought result to be adjusted, indicated by the input question information in the first intermediate thought content, includes: inserting at least one thought adjustment node into the thought processing information. The thought adjustment node represents the thought step to be adjusted, which is the thought step indicated by the input question information. This will be discussed in conjunction with... Figure 4C Please provide an explanation.

[0082] Figure 4C This is a schematic diagram illustrating the adjusted thinking result obtained according to an embodiment of the present disclosure.

[0083] like Figure 4C As shown, the visual interface i400 can also display third thinking process information g403. This third thinking process information g403 can be obtained after inserting a thinking adjustment node into the second thinking process information g402. The aforementioned intent recognition result int401 includes the adjustment instruction information "replan action plan". Therefore, it can also be determined that the thinking step indicated by the input question information can be a skill planning step. The thinking adjustment node can be, for example,... Figure 4C The redesigned nodes are shown.

[0084] After inserting a thinking adjustment node, the thinking steps to be adjusted represented by the thinking adjustment node can be executed to adjust the thinking results.

[0085] In some embodiments, adjusting at least one thought result to be adjusted, indicated by the input question information in the first intermediate thought content, includes: generating a sub-result to be added for the thought result to be adjusted based on the intent recognition result. The sub-result to be added is then added to the position indicated by the intent recognition result in the thought result to be adjusted, resulting in the adjusted thought result. For example, based on the information "add steps to generate practice questions" added to the intent recognition result in int401, a sub-result to be added, "Write practice questions: I will organize and write 3 sets of practice questions for you based on the previously summarized key knowledge points," can be generated. Next, this sub-result to be added can be added to the position indicated by the intent recognition result, resulting in the adjusted thought result t412. This position can be after the intermediate thought sub-result "Create demonstration animation: You mentioned demonstration animation in your request, presumably you also hope to allow students to concretely experience abstract knowledge in a more intuitive and engaging way. I will call on code experts to design the scene and create the animation." Through this embodiment of the disclosure, a sub-result to be added is generated based on the input question information, effectively preserving the generated thought content, allowing the agent or large model to continue thinking quickly and accurately, and effectively utilizing the generated context information.

[0086] In some embodiments, adjusting at least one thought result to be adjusted, as indicated by the input question information in the first intermediate thought content, may further include: displaying the adjusted thought result in the visible area of ​​the thought adjustment node in a visual interface. For example... Figure 4C As shown, the adjusted thinking result t412 can be displayed in the visible area used for replanning nodes. The second intermediate thinking content can include this adjusted thinking result t412.

[0087] It is understandable that the above example of adding sub-results with low relevance to intermediate thinking sub-results serves as an illustration of this disclosure. However, this disclosure is not limited to this; it is also possible to add sub-results with high relevance to intermediate thinking sub-results. For example, if the follow-up question is "The filename should be 'Chapter 4'", a sub-result to be added to the above-mentioned thinking results can be "Continue searching for the lecture outline: I will carefully search for documents with filenames containing 'Chapter 4' in the cloud drive location you specify." Thus, the thinking content can be corrected if the generated thinking content is incomplete or inaccurate.

[0088] It can also be understood that the above description uses the addition of one or more thought sub-results as an example to illustrate this disclosure. However, this disclosure is not limited to this. In some other embodiments, based on the intent recognition result, at least one intermediate thought sub-result in the thought result to be adjusted can be modified to obtain the adjusted thought result. For example, if the agent's generated thought content can be modified by follow-up information, based on the above follow-up information "the file name should be 'Chapter 4'", the intermediate thought sub-result "Looking for lecture outline: I will carefully search for documents with file names containing 'friction', 'inclined plane motion', 'teaching notes' in the cloud drive location you specify" can be modified to "Looking for lecture outline: I will carefully search for documents with file names containing 'Chapter 4' in the cloud drive location you specify". This makes the adjusted thought result more concise and effectively improves the user experience.

[0089] As can be understood, the above text has explained the method for determining the second intermediate consideration content. The following text will explain some of the methods for generating response information in this disclosure.

[0090] In some embodiments of operation S230, generating response information for at least one of the initial question information and input information based on the second intermediate thinking content includes: continuing to execute at least one thought step to be executed in the thinking process based on the second intermediate thinking content to obtain response information. The thought step to be executed is a thought step that was not completed in the thinking process when the first intermediate thinking content was saved. For example, during the execution of the skill planning step, the first part of the thinking result t411 was obtained as the first intermediate thinking content. After determining the adjusted thinking result t412, the skill planning step can continue to be executed. After the skill planning step is completed, the "planning execution" step and the "thinking completion" step can be executed sequentially to obtain response information. The response information can be the result for the initial question information and follow-up question information. Through the embodiments of this disclosure, after adjusting the first intermediate thinking content using input information to obtain the second intermediate thinking content, it is possible to continue thinking along a thinking direction that is more in line with the user's expectations while retaining most of the context information, which can effectively save the hardware resources required for information processing and improve processing efficiency.

[0091] It is understood that the above example of inputting follow-up information was used to illustrate this disclosure. However, this disclosure is not limited to this; the input information can also be editable information, as will be explained below.

[0092] Figures 5A to 5D This is a schematic diagram of another visual interface according to this disclosure.

[0093] like Figure 5A As shown, the visual interface i500 can display the initial problem information q501 and the input box ib40. It is understood that the above descriptions of the visual interface i300, the initial problem information q301, and the input box ib30 also apply to the visual interface i500, the initial problem information q501, and the input box ib40, and will not be repeated here.

[0094] like Figure 5A As shown, when executing the "Planning Execution" step, a presentation file can be generated based on the generated presentation outline. The visual interface i500 also displays the second part of the thinking results t521 for the "Planning Execution" step. If the pause control is triggered, the saved first intermediate thinking content can include the second part of the thinking results t521 and the intermediate thinking results of the skill planning step. The second part of the thinking results t521 can include the presentation outline. The titles of each chapter in the presentation outline can be used as editable material.

[0095] The visual interface i500 may also include fourth-order processing information g504. This fourth-order processing information g504 may include planning execution nodes. The visual interface i500 may also display editing controls r501 in the visible area corresponding to the planning execution nodes. For example... Figure 5B As shown, after the user triggers the editing control r501, the visual interface i500 can display the presentation outline for editing. For example, the user can select... Figure 5B The title of Section 2.1 is "Key Content Extraction," and the input is "Example Exercises." It can be understood that the title "Key Content Extraction" can be editable material related to the aforementioned objectives. "Example Exercises" can be user-inputted and can be used as the editable input information mentioned above.

[0096] After modifying the presentation outline, the user can trigger the edit completion control r502 to obtain second intermediate thinking content. The second intermediate thinking content may include at least one edited material. The edited material can be obtained by editing the target editable material using input editing information. For example, after editing the target editable material "User Needs Extraction" using the input editing information "Practice Examples," the resulting edited material can be that input editing information. Through this embodiment of the disclosure, based on input editing information, the generated thinking content can be adjusted quickly and efficiently, effectively improving information processing efficiency, interaction efficiency, and user experience.

[0097] In some embodiments, in other implementations of the above-described operation S220, determining the second intermediate thought content may further include: in thinking

[0098] Insert a content modification node in the diagram. A content modification node represents a second, newly added step that determines a second intermediate thought process based on at least one piece of input editing information. For example... Figure 5C As shown, a modification node can be added to the fourth thinking process information g504 to obtain the fifth thinking process information g505. This modification node can be used as a content modification node. In the visible area of ​​this content modification node, the preset text "You modified my answer, and I will make adjustments based on your modification" can be displayed.

[0099] Next, operation S230 can be executed to continue processing at least one pending thought step based on the second intermediate thought content, and obtain response information. For example... Figure 5D As shown, you can continue with the "Plan Execution" step.

[0100] It is understood that the above description, using the example of replacing the target editable material with input editing information, illustrates this disclosure. However, this disclosure is not limited to this; the input editing information can instruct various forms of editing of the target editable material. For example, the input editing information can also instruct the deletion of one or more selected characters to change "key content extraction" to "key content," thereby displaying the original text of the key content in the generated presentation. As another example, the input editing information can also instruct the addition of one or more input characters to change "key content extraction" to "key content extraction of similar chapters," thereby displaying the key content extraction results of similar chapters in the generated presentation. Similar chapters may, for example, include chapters related to other mechanics knowledge.

[0101] It is understood that the above description uses examples of input question information and input editing information to illustrate this disclosure. However, this disclosure is not limited to this; the input information may include at least one input question information and at least one input editing information to allow for multiple adjustments to the first intermediate thought content.

[0102] It can also be understood that the above description of the present disclosure uses the example of the first intermediate thinking content including at least one of intermediate thinking results and partial thinking results. However, the present disclosure is not limited to this. In some embodiments, the initial question information and input information can be provided by at least one target object. The first intermediate thinking content can also include object description information for the target object. Taking a user as an example, the object description information can include user profiles, preferences, and other information. For example, the target object can also be other intelligent agents or large models. Through the embodiments of the present disclosure, the first intermediate thinking content can also include object description information, which can effectively retain richer information, save the hardware and time costs required to reacquire object description information when continuing to think, and effectively improve information processing efficiency. It can be understood that the acquisition, processing, and use of object description information comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0103] In some embodiments, generating a response result for at least one of the initial question information and input information based on the second intermediate thinking content includes: generating a response result displayed on a visual interface based on the second intermediate thinking content. For example, the response result may also be displayed on the aforementioned visual interface.

[0104] It is understood that the above operation S220 may further include: in response to detecting input information and determining that the correlation index value between the input information and the initial question information is greater than or equal to a preset correlation threshold, determining the second intermediate thinking content based on the first intermediate thinking content and the input information. Therefore, after executing operation S230, a response result can be generated for the input information and the initial question information.

[0105] In other embodiments, the method may further include: in response to determining that the correlation index value between the input information and the initial question information is less than a preset correlation threshold, a response result for the input information can be generated.

[0106] It is understood that the above description uses a large model performing thought processing as an example to illustrate this disclosure. However, this disclosure is not limited to this. The target object can provide question information and input information to an intermediate agent. This intermediate agent performs thought processing and provides the result of the thought processing to the large model so that the large model can generate response information. It can also be understood that after the above "thought completion" step is completed, response information can be output, or response information can be generated based on the processing result of the thought processing. The large model can be a large language model (LLM), a conversational large model, etc.

[0107] As can be understood, the method of this disclosure has been described above, and the apparatus of this disclosure will be described below.

[0108] Figure 6 This is a block diagram of an information processing apparatus according to an embodiment of the present disclosure.

[0109] like Figure 6 As shown, the device 600 may include a detection module 610, a determination module 620, and a generation module 630.

[0110] The detection module is used to detect input information during the thinking process based on the initial question information when interacting with a large model.

[0111] The module 620 is used to determine the second intermediate thinking content based on the first intermediate thinking content information processing and input information. The first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information;

[0112] The generation module 630 is used to generate response information based on the second intermediate thinking content, targeting at least one of the initial problem information and input information for information processing.

[0113] In some embodiments, the device 600 further includes: a pause module, configured to pause the thinking process based on the initial question information in response to receiving at least one of a pause signal and input information via a visual interface; and a save module, configured to save the thinking content generated during the thinking process based on the initial question information as first intermediate thinking content.

[0114] In some embodiments, the first intermediate thought content is displayed on the visual interface, and the pause signal is issued after the pause control displayed on the visual interface is triggered.

[0115] In some embodiments, the first intermediate thinking content is displayed on a visual interface, the thinking process includes at least one thinking step, the first intermediate thinking content includes at least one intermediate thinking result and at least one partial thinking result, at least one of the intermediate thinking result and the partial thinking result includes editable material, the intermediate thinking result is the result obtained after completing a thinking step, and the partial thinking result is the result obtained during the execution of the thinking step.

[0116] In some embodiments, the input information includes at least one of the following received during the thinking process: input editing information for a target editable material; and input question information. The target editable material is a selected editable material from at least one editable material. The input question information is received via an input module displayed on a visual interface, and the input question information is used to indicate at least one thinking result to be adjusted in the first intermediate thinking content.

[0117] In some embodiments, the determining module 620 includes: a first determining submodule, configured to determine the intent recognition result of the input question information; and an adjusting submodule, configured to adjust at least one thought result to be adjusted indicated by the input question information in the first intermediate thinking content according to the intent recognition result, to obtain a second intermediate thinking content. The second intermediate thinking content includes: the adjusted thinking result obtained by adjusting the thought result to be adjusted.

[0118] In some embodiments, intermediate thinking results and partial thinking results each include at least one intermediate thinking sub-result. The adjustment submodule includes at least one of the following: an adding subunit, configured to generate a sub-result to be added for the thinking result to be adjusted based on the intent recognition result, and add the sub-result to be added to the thinking result to be adjusted at the position indicated by the intent recognition result, to obtain the adjusted thinking result; and a modifying subunit, configured to modify at least one intermediate thinking sub-result in the thinking result to be adjusted based on the intent recognition result, to obtain the adjusted thinking result.

[0119] In some embodiments, the second intermediate consideration includes at least one edited material, which is obtained by editing the target editable material using input editing information.

[0120] In some embodiments, the visual interface includes thinking process information, which includes at least one thinking node, representing a thinking step. The first determining submodule includes: a first insertion unit, configured to insert a problem analysis input problem node into the thinking process information, the problem analysis input problem node representing a first new step in intent recognition of the input problem information; and a first display unit, configured to display the intent recognition result in a visible area of ​​the visual interface corresponding to the input problem node.

[0121] In some embodiments, the adjustment submodule includes: a second insertion unit, configured to insert at least one thinking adjustment node into the thinking process information, wherein the thinking adjustment node represents the thinking step to be adjusted, and the thinking step to be adjusted is the thinking step indicated by the input question information; and a second display unit, configured to display the adjusted thinking result in the visible area of ​​the thinking adjustment node in the visual interface.

[0122] In some embodiments, the visual interface includes thinking process information, which includes at least one thinking node representing a thinking step. The determining module 620 includes an insertion submodule for inserting a content modification node into the thinking process information, the content modification node representing a second new step in determining second intermediate thinking content based on at least one input editing information.

[0123] In some embodiments, the generation module 630 includes an execution submodule, configured to continue executing at least one thought step to be executed in the thought process based on the second intermediate thought content, and obtain response information, wherein the thought step to be executed is a thought step that was not completed in the thought process when the first intermediate thought content was saved.

[0124] In some embodiments, the initial question information and input information are provided by at least one target object, and the first intermediate thinking content also includes object description information for the target object.

[0125] In some embodiments, the generation module 630 includes a generation submodule for generating response information to be displayed on a visual interface based on the second intermediate thought content.

[0126] It is understood that the apparatus of this disclosure has been described above, and the intelligent agent of this disclosure will be described below.

[0127] Figure 7 This is a schematic diagram of an intelligent agent according to an embodiment of the present disclosure.

[0128] In embodiments of this disclosure, the von Neumann architecture in modern computer theory is inspired, such as... Figure 7 As shown, the intelligent agent 700 may include multiple core modules: a receiving module 710, a processing module 720, and an output module 730.

[0129] In the example, the receiving module 710 is responsible for receiving or sensing information such as queries, requests, instructions, signals, question information, or data from the outside world (e.g., users or the external environment), and converting it into a format that the intelligent agent 700 can understand and process. The receiving module 710 is the primary link in the interaction between the intelligent agent 700 and the outside world, enabling the intelligent agent 700 to efficiently and accurately acquire necessary "sensory" information from the outside world and respond to this information. In the example, the receiving module 710 can be used to receive information to be processed. This information may include initial question information and input information as described above.

[0130] In the example, processing module 720 is the core support for the ability of agent 700 to handle complex tasks. Processing module 720 executes method 200 described above.

[0131] In the example, output module 730 can be used to output the output information obtained by processing module 720. The output information may include response information as described above.

[0132] In the example, the processing module 720 may include a control unit 721, a storage unit 722, and a processing unit 723.

[0133] During operation, the control unit 721 will continuously interact with the storage unit 722, the arithmetic unit 723, and / or the output module 730. However, in the embodiments of this disclosure, the control unit 721 initiates communication with the storage unit 730, the arithmetic unit 723, and / or the output module 730 as a single initiator, while there may be no communication coupling between the storage unit 722, the arithmetic unit 723, and the output module 730.

[0134] In the example, the performance of control unit 721 is closely related to the large model on which agent 700 is based. To fully leverage the capabilities of the large language model, the internal structure of control unit 721 can be designed to be highly configurable and scalable to handle various types of tasks and requirements in real-world scenarios.

[0135] Storage unit 722 can be responsible for remembering information such as historical dialogues and event streams. Various intermediate thought processes and user profile data, as described above, can be included in storage unit 722.

[0136] In the example, after acquiring the information to be processed, the agent 700 can retrieve at least one retrieval result for the information to be processed from the storage unit 722 and feed it back to the control unit 721.

[0137] The arithmetic unit 723 can be viewed as a predefined tool library. Tools such as format conversion tools can be included in the arithmetic unit 723.

[0138] In the example, when the AI ​​agent 700 needs to render multiple output data, it can call the relevant renderers and display tools from the computing unit 723 and feed them back to the processing module 720. Then, the processing module 720 can use the returned renderers and display tools to render the results and pass them to the output module 730. It's understandable that although large language models have excellent language understanding and generation capabilities, like humans, the tasks they can solve without any tools are very limited. When the AI ​​agent 700 is given the ability to call tools, it can perform tasks such as displaying results.

[0139] The intelligent agent 700 according to the embodiments of this disclosure can simply and effectively improve the level of intelligence, and enhance flexibility and versatility.

[0140] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0141] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0142] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0143] like Figure 8 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0144] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0145] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as information processing methods. For example, in some embodiments, the information processing method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the information processing method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform an information processing method by any other suitable means (e.g., by means of firmware).

[0146] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard parts (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0148] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory (EPROM) or flash memory, optical fiber, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0149] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a cathode ray tube (CRT) monitor or a liquid crystal display (LCD)); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0150] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0151] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0152] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0153] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. An information processing method, comprising: During a round of interaction with the large model, the input information is detected during the thinking process based on the initial problem information; The input information includes text information; In response to the detection of the input information, a second intermediate thinking content is determined based on the first intermediate thinking content and the input information. The first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information. Based on the second intermediate thinking content, generate response information for the initial question information and the input information, or generate response information for the input information. The input information includes input question information, which is used to indicate at least one thought result to be adjusted in the first intermediate thinking content. The input question information is received via an input module displayed on the visual interface. The determination of the second intermediate consideration content includes: Determine the intent recognition result of the input question information; Based on the intent recognition result, at least one of the thought results to be adjusted indicated by the input question information in the first intermediate thinking content is adjusted to obtain the second intermediate thinking content, which includes: the adjusted thinking result obtained by adjusting the thought result to be adjusted.

2. The method according to claim 1, further comprising: In response to receiving at least one of a pause command and the input information via a visual interface, the thought process based on the initial question information is paused; Save the thought content generated during the thought process based on the initial problem information as the first intermediate thought content.

3. The method according to claim 1, wherein, The first intermediate thinking content is displayed on the visual interface. The thinking process includes at least one thinking step. The first intermediate thinking content includes at least one intermediate thinking result and at least one partial thinking result. At least one of the intermediate thinking result and the partial thinking result includes editable material. The intermediate thinking result is the result obtained after completing one of the thinking steps. The partial thinking result is the result obtained during the execution of the thinking step.

4. The method according to claim 3, wherein, The input information also includes input editing information for the target editable material received during the execution of the thinking process, wherein the target editable material is at least one of the editable materials selected.

5. The method according to claim 4, wherein, The intermediate thinking results and the partial thinking results each include at least one intermediate thinking sub-result. The adjustment of at least one of the following thinking results, which involves adjusting at least one of the input question information indications in the first intermediate thinking content: Based on the intent recognition result, a sub-result to be added to the thought result to be adjusted is generated, and the sub-result to be added to the thought result to be adjusted is added to the position indicated by the intent recognition result to obtain the adjusted thought result; Based on the intent recognition result, at least one of the intermediate thinking sub-results in the thinking result to be adjusted is modified to obtain the adjusted thinking result.

6. The method according to claim 4, wherein, The second intermediate consideration includes at least one edited material, which is obtained by editing the target editable material using the input editing information.

7. The method according to claim 4, wherein, The visual interface includes thinking process information, which includes at least one thinking node, used to represent the thinking step. The intent recognition result for determining the input question information includes: An input question node is inserted into the thought process information, and the input question node is used to represent the first new step of intention recognition of the input question information; The intent recognition result is displayed in the visible area of ​​the input question node in the visual interface.

8. The method according to claim 7, wherein, The adjustment of at least one of the thinking results to be adjusted in the first intermediate thinking content includes: Insert at least one thinking adjustment node into the thinking process information. The thinking adjustment node is used to represent the thinking step to be adjusted, which is the thinking step indicated by the input question information. The adjusted thinking results are displayed in the visible area of ​​the thinking adjustment node in the visual interface.

9. The method according to claim 4, wherein, The visual interface includes thinking process information, which includes at least one thinking node, used to represent the thinking step. The determination of the second intermediate consideration content includes: A content modification node is inserted into the thinking process information, the content modification node being used to represent a second new step of determining the second intermediate thinking content based on at least one of the input editing information.

10. The method according to claim 3, wherein, The step of generating response information based on the second intermediate thinking content for at least one of the initial question information and the input information includes: Based on the second intermediate thinking content, at least one pending thinking step of the thinking process is executed to obtain the response information. The pending thinking step is a thinking step that was not completed in the thinking process when the first intermediate thinking content was saved.

11. The method according to claim 2, wherein, The initial question information and the input information are provided by at least one target object, and the first intermediate thinking content also includes object description information for the target object.

12. The method according to claim 1, wherein, The step of generating response information based on the second intermediate thinking content for at least one of the initial question information and the input information includes: Based on the second intermediate thought content, the response information displayed on the visual interface is generated.

13. An information processing method, comprising: During a round of interaction with the large model, the input information is detected during the thinking process based on the initial problem information; The input information includes text information; In response to the detection of the input information, a second intermediate thinking content is determined based on the first intermediate thinking content and the input information. The first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information. Based on the second intermediate thinking content, generate response information for the initial question information and the input information, or generate response information for the input information. The first intermediate thought process is displayed on the visual interface. The input information includes editing information for the target editable material, which is at least one editable material selected from among at least one editable material. The visual interface includes thinking process information, which includes at least one thinking node, used to represent the thinking step. The determination of the second intermediate consideration content includes: A content modification node is inserted into the thinking process information, the content modification node being used to represent a second new step of determining the second intermediate thinking content based on at least one of the input editing information.

14. The method of claim 13, further comprising: In response to receiving at least one of a pause command and the input information via a visual interface, the thought process based on the initial question information is paused; Save the thought content generated during the thought process based on the initial problem information as the first intermediate thought content.

15. The method according to claim 13, wherein, The thinking process includes at least one thinking step. The first intermediate thinking content includes at least one of at least one intermediate thinking result and at least one of the partial thinking results. At least one of the intermediate thinking result and the partial thinking result includes editable material. The intermediate thinking result is the result obtained after completing one of the thinking steps, and the partial thinking result is the result obtained during the execution of the thinking step.

16. The method according to claim 15, wherein, The input information also includes input question information received during the execution of the thinking process; The input question information is received via an input module displayed on the visual interface, and the input question information is used to indicate at least one thought result to be adjusted in the first intermediate thought content.

17. The method according to claim 16, wherein, The determination of the second intermediate consideration content includes: Determine the intent recognition result of the input question information; Based on the intent recognition result, at least one of the thought results to be adjusted indicated by the input question information in the first intermediate thinking content is adjusted to obtain the second intermediate thinking content, which includes: the adjusted thinking result obtained by adjusting the thought result to be adjusted.

18. The method according to claim 17, wherein, The intermediate thinking results and the partial thinking results each include at least one intermediate thinking sub-result. The adjustment of at least one of the following thinking results, which involves adjusting at least one of the input question information indications in the first intermediate thinking content: Based on the intent recognition result, a sub-result to be added to the thought result to be adjusted is generated, and the sub-result to be added to the thought result to be adjusted is added to the position indicated by the intent recognition result to obtain the adjusted thought result; Based on the intent recognition result, at least one of the intermediate thinking sub-results in the thinking result to be adjusted is modified to obtain the adjusted thinking result.

19. The method of claim 16, wherein, The second intermediate consideration includes at least one edited material, which is obtained by editing the target editable material using the input editing information.

20. The method of claim 17, wherein, The visual interface includes thinking process information, which includes at least one thinking node, used to represent the thinking step. The intent recognition result for determining the input question information includes: An input question node is inserted into the thought process information, and the input question node is used to represent the first new step of intention recognition of the input question information; The intent recognition result is displayed in the visible area of ​​the input question node in the visual interface.

21. The method according to claim 20, wherein, The adjustment of at least one of the thinking results to be adjusted in the first intermediate thinking content includes: Insert at least one thinking adjustment node into the thinking process information. The thinking adjustment node is used to represent the thinking step to be adjusted, which is the thinking step indicated by the input question information. The adjusted thinking results are displayed in the visible area of ​​the thinking adjustment node in the visual interface.

22. The method according to claim 15, wherein, The step of generating response information based on the second intermediate thinking content for at least one of the initial question information and the input information includes: Based on the second intermediate thinking content, at least one pending thinking step of the thinking process is executed to obtain the response information. The pending thinking step is a thinking step that was not completed in the thinking process when the first intermediate thinking content was saved.

23. The method according to claim 14, wherein, The initial question information and the input information are provided by at least one target object, and the first intermediate thinking content also includes object description information for the target object.

24. The method according to claim 13, wherein, The step of generating response information based on the second intermediate thinking content for at least one of the initial question information and the input information includes: Based on the second intermediate thought content, the response information displayed on the visual interface is generated.

25. An information processing apparatus, comprising: The detection module is used to detect the input information during the thinking process based on the initial question information during a round of interaction with the large model; The input information includes text information; The determining module is used to determine the second intermediate thinking content based on the first intermediate thinking content and the input information. The first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information. The generation module is used to generate response information for the initial question information and the input information, or to generate response information for the input information, based on the second intermediate thinking content. The input information includes input question information, which is used to indicate at least one thought result to be adjusted in the first intermediate thought content; The input question information is received via an input module displayed on the visual interface; The determining module includes: The first determining submodule is used to determine the intent recognition result of the input question information; The adjustment submodule is used to adjust at least one thought result to be adjusted indicated by the input question information in the first intermediate thought content according to the intent recognition result, so as to obtain the second intermediate thought content. The second intermediate thought content includes: the adjusted thought result obtained by adjusting the thought result to be adjusted.

26. An information processing apparatus, comprising: The detection module is used to detect the input information during the thinking process based on the initial question information during a round of interaction with the large model; The input information includes text information; The determining module is used to determine the second intermediate thinking content based on the first intermediate thinking content and the input information. The first intermediate thinking content is the thinking content generated during the thinking process based on the initial question information. The generation module is configured to generate response information based on the second intermediate thinking content, targeting at least one of the initial question information and the input information. The first intermediate thought process is displayed on the visual interface. The input information includes editing information for the target editable material, which is at least one selected editable material among the editable materials. The visual interface includes thinking process information, which includes at least one thinking node, and the thinking node is used to represent thinking steps. The determining module includes: An insertion submodule is used to insert a content modification node into the thinking process information, the content modification node being used to represent a second addition step in determining the second intermediate thinking content based on at least one input editing information.

27. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 24.

28. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 24.

29. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 24.

Images