Information processing method, information processing device, and computer program

Abstract

Provided are an information processing method, an information processing device, and a computer program.　The present invention uses a computer to perform a process for: receiving device information and / or the purpose of use of a substrate processing device; generating a response text according to the received device information and / or purpose of use, by using a language model that refers to an output dataset obtained by performing a simulation related to substrate processing; and outputting the generated response text.

Description

Information Processing Method, Information Processing Apparatus, and Computer Program

[0001] The present invention relates to an information processing method, an information processing apparatus, and a computer program.

[0002] In the development of a substrate processing apparatus, various simulations can output physical quantities and spatio-temporal information that cannot be measured only by sensors, and are used for recipe development, chamber design, identification of causes of equipment troubles, and the like.

[0003] In order to perform simulations related to a substrate processing apparatus, it is necessary to concatenate multiple types of simulations due to differences in the scale of the computable region. Furthermore, even for simulations of the same level, there are a wide variety from approximate models to first-principles models. Therefore, it is necessary for the developer to select a model for the required problem and perform operations from the creation, analysis, and interpretation of the input information of each model.

[0004] Japanese Patent Application Laid-Open No. 2018-125451

[0005] The present disclosure aims to provide an information processing method, an information processing apparatus, and a computer program that can present information (reply text) based on the execution results of simulations performed in the past to a developer according to the apparatus information and / or usage purpose of a substrate processing apparatus.

[0006] The information processing method of the present disclosure receives the apparatus information and / or usage purpose of a substrate processing apparatus, generates a reply text corresponding to the received apparatus information and / or usage purpose using a language model that refers to an output data group obtained by executing a simulation related to substrate processing, and causes a computer to execute a process of outputting the generated reply text.

[0007] According to the present disclosure, information (reply text) based on the execution results of simulations performed in the past can be presented to a developer according to the apparatus information and / or usage purpose of a substrate processing apparatus.

[0008] This is a schematic diagram showing the configuration of an information processing system according to an embodiment. This is a conceptual diagram showing an example of a database. This is a block diagram showing the internal configuration of a developer terminal. This is an explanatory diagram illustrating an example of a response output to a developer query. This is an explanatory diagram illustrating an example of a response output to a developer query. This is an explanatory diagram illustrating an example of a response output to a developer query. This is an explanatory diagram illustrating an example of a response output to a developer query. This is an explanatory diagram illustrating an example of a response output to a developer query. This is a flowchart illustrating the procedure of processing executed by the developer terminal.

[0009] An embodiment will be described below with reference to the drawings. (Embodiment 1) Figure 1 is a schematic diagram showing the configuration of an information processing system according to the embodiment. The information processing system according to the embodiment receives device information and / or purpose of use of a substrate processing apparatus as a query from a developer who is developing a process for substrate processing. The substrate processing system may accept either the device information and the purpose of use of the substrate processing apparatus as a query, or it may accept both the device information and the purpose of use. The information processing system generates a response sentence using a language model that refers to the output data set obtained by executing a simulation related to substrate processing, and provides the generated response sentence to the developer.

[0010] The information processing system includes a developer terminal 100, a circuit board processing unit 120, a device data infrastructure 140, storage 160, and the like. These developer terminal 100, circuit board processing unit 120, device data infrastructure 140, and storage 160 are connected to each other via a local communication network NW1, such as a LAN (Local Area Network) or dedicated line, of the company, university, or research institution to which the developer belongs.

[0011] The substrate processing apparatus 120 is a device for performing substrate processing on a substrate (wafer) to be processed, such as etching, film deposition, CMP (Chemical Mechanical Polishing), ashing, and cleaning. Alternatively, the substrate processing apparatus 120 may be a device for manufacturing FPDs (Flat Panel Displays) such as liquid crystal display panels and organic EL (Electro-Luminescence) panels. The information processing system illustrated in Figure 1 is configured with one substrate processing apparatus 120, but it may also be configured with multiple substrate processing apparatuses 120. Furthermore, the substrate processing apparatus 120 may be a group of apparatuses (substrate processing system) composed of multiple types of substrate processing apparatuses.

[0012] The developer terminal 100 is a terminal device such as a personal computer used by the developer. The developer performs process development by running various simulations related to substrate processing in the substrate processing device 120 using the developer terminal 100. Various simulation models (hereinafter also simply referred to as models) used in the simulation may be installed on the developer terminal 100, or they may be installed on a server device (not shown) that can be accessed from the developer terminal 100.

[0013] The models include, for example, models that simulate gas-phase reactions in substrate processing, models that simulate transport phenomena, and models that simulate the shape of the resulting product. Models that simulate gas-phase reactions include gas-phase reaction models, fluid models, particle models, machine learning models, and surrogate models. Models that simulate transport phenomena include circuit models, sheath particle models, kinetic models, machine learning models, and surrogate models. Models that simulate shape include probabilistic shape models, surface reaction models, machine learning models, and surrogate models.

[0014] Developers may use a single model to simulate a specific phenomenon, or they may combine multiple models to perform simulations that predict the state and outcome within the chamber. For example, when performing a simulation of wafer processing shape, developers can select a combination of multiple models that simulate gas-phase reactions, transport phenomena, and shape, and then provide parameters such as equipment information and recipe information to simulate the wafer processing shape.

[0015] The device data infrastructure 140 is a group of systems for collecting, managing, and analyzing various data related to substrate processing. In this embodiment, the device data infrastructure 140 includes a database 141 that stores various data (hereinafter referred to as model data) obtained by executing simulations related to substrate processing. When a simulation related to substrate processing is executed on the developer terminal 100, the device data infrastructure 140 collects data from the developer terminal 100, such as device information of the substrate processing apparatus 120 that was the target of the simulation, information on the process and recipe in which the simulation is executed, information on the selected simulation model, the reason for selecting the model, and the results of the simulation, and stores the collected data as model data in the database 141.

[0016] Figure 2 is a conceptual diagram showing an example of database 141. Database 141 stores associated information such as equipment name, process name, recipe, model, reason for model selection, and execution results. The equipment name represents the name of the equipment to be simulated (for example, the substrate processing device 120). The process name represents the name of the process to be simulated. A recipe represents a procedure designed in advance by a developer or the like to realize a particular process. Each recipe consists of multiple steps. Each step represents the smallest processing unit that changes the state (attributes of the object to be processed or the state of the substrate) in the substrate processing process.

[0017] The "Model" field represents the name of the simulation model used in the simulation. As described above, simulations related to substrate processing utilize various simulation models, including those for simulating gas-phase reactions, transport phenomena, and shape simulations. Developers select simulation models based on factors such as accuracy, computation speed, computation cost, and linearity / nonlinearity, and then execute simulations related to substrate processing using the selected models. Database 141 stores not only the simulation models selected by the developers but also the reasons for their selection. In the example in Figure 2, for apparatus A001, process a001, and recipe α001, a particle model with high nonlinearity was selected to simulate gas-phase reactions. Furthermore, for apparatus B001, process b001, and recipe β001, a fluid model, kinetic model, and stochastic shape model were selected based on computation cost to simulate gas-phase reactions, transport phenomena, and shape.

[0018] The execution results represent data obtained as a result of running the simulation. The execution results may be numerical data consisting of multiple numerical values, or they may be image data, which is a graph or image of the numerical values ​​obtained as a result of the simulation. The experimental results registered in database 141 may be links to files containing numerical data or image data.

[0019] The device data board 140 stores various data related to the substrate processing device 120, in addition to model data obtained by running simulations.

[0020] For example, the substrate processing apparatus 120 is equipped with parts such as a heater, gas supply source, flow rate controller, and high-frequency power supply, and the apparatus has information about these parts as device information. When substrate processing is performed in the substrate processing apparatus 120, recipe information that defines the setting values ​​for the above parts, processing procedures, etc., is input. While substrate processing is being performed in the substrate processing apparatus 120 according to the recipe information, the temperature of the substrate, the pressure and gas flow rate in the chamber, the voltage applied to the upper and lower electrodes, the plasma emission intensity, etc. are measured and output as measurement data. In addition, after the completion of substrate processing, various measurements of the substrate obtained as a result (for example, shape measurement) may be performed.

[0021] The device data platform 140 collects device information of the substrate processing apparatus 120, information on the processes and recipes used for substrate processing, measurement data measured during the execution of substrate processing, and measurement data measured on the output products after substrate processing from the developer terminal 100 or the substrate processing apparatus 120, and stores the collected data.

[0022] Storage 160 is a device that stores data from various documents related to circuit board processing. The document data stored in Storage 160 includes document manuals for various simulation models and script collections used in simulators. The simulation model document manuals are data that document the purpose, usage, input / output parameters, etc. of the simulation model, and are created by the developer or an external expert. The script collections are document data that include the purpose and usage of various commands, parameters to be given to commands, example sentences, etc., and are created by the developer or an external expert. It is assumed that Storage 160 stores such document data.

[0023] Storage 160 may store not only data on literature related to substrate processing simulations, but also data on documents created by process development experts, such as reports on substrate processing, internal documents, research logs, and academic papers. Experts may be the developers themselves, or engineers, analysts, researchers, etc., other than developers. Documents created by experts may include not only text, but also mathematical formulas, chemical reaction equations, figures, tables, graphs, flowcharts, etc. The literature data stored in storage 160 is local data accessible via the communication network NW1, and access from outside the communication network NW1 may be restricted.

[0024] The information processing system further includes an LLM server 200, storage 220, etc. These LLM server 200 and storage 220 are connected via a communication network NW2, such as the Internet.

[0025] The LLM server 200 is a server computer that provides services using a Large Language Model (LLM). The LLM server 200 is equipped with a language model MD that generates response sentences to input queries. The language model MD is an existing large language model such as GPT-4 (Generative Pretrained Transformer 4), LLaMA (Large Language Model Meta AI), or BERT (Bidirectional Encoder Representations from Transformers). Alternatively, the language model MD may be a proprietary language model that has been tuned from an existing large language model. The language model MD may be a unimodal language model that outputs a response sentence in text data in response to a question sentence in text data, or it may be a multimodal language model that can accept inputs other than text data, such as image data or audio data.

[0026] Storage 220 is a device for storing data such as specialized literature related to substrate processing (particularly substrate processing simulation). The data stored in storage 220 is the same as the literature data stored in storage 160, except that it is made publicly available via the communication network NW2. The data stored in storage 220 consists of documents created by process development experts and includes various simulation model document manuals, script collections used in simulators, specialized books, and academic papers published on the communication network NW2 by companies, universities, research institutions, etc.

[0027] The developer terminal 100 receives device information and / or intended use of the substrate processing device 120 as a query from the developer, and sends the received query to the LLM server 200. The terminal then generates a response using the language model MD, which refers to the output data obtained by executing a simulation related to substrate processing, and outputs the generated response.

[0028] Figure 3 is a block diagram showing the internal configuration of the developer terminal 100. The developer terminal 100 is a dedicated or general-purpose computer comprising, for example, a control unit 101, a storage unit 102, a communication unit 103, an operation unit 104, and a display unit 105.

[0029] The control unit 101 includes a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), and the like. The ROM in the control unit 101 stores control programs that control the operation of each hardware component of the developer terminal 100. The CPU in the control unit 101 reads and executes the control programs stored in the ROM and the computer programs described later stored in the memory unit 102, and controls the operation of each hardware component, thereby making the entire device function as an information processing device according to the present disclosure. The RAM in the control unit 101 temporarily stores data used during the execution of calculations.

[0030] In this embodiment, the control unit 101 is configured to include a CPU, ROM, and RAM, but the configuration of the control unit 101 is not limited to the above. The control unit 101 may be one or more control circuits or arithmetic circuits equipped with, for example, a GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), quantum processor, volatile or non-volatile memory, etc. Furthermore, the control unit 101 may include functions such as a clock that outputs date and time information, a timer that measures the elapsed time from the time a measurement start instruction is given to the time a measurement end instruction is given, and a counter that counts numbers.

[0031] The storage unit 102 includes storage devices such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), and an EEPROM (Electronically Erasable Programmable Read Only Memory). The storage unit 102 stores various computer programs executed by the control unit 101 and various data used by the control unit 101.

[0032] The computer program (program product) stored in the memory unit 102 includes an answer generation program PG that receives device information and / or purpose of use of the substrate processing apparatus 120 as a query, generates an answer statement corresponding to the received query using a language model MD that refers to a group of output data (model data) obtained by executing a simulation related to substrate processing, and causes the computer to execute a process to output the generated answer statement. The answer generation program PG may be a single computer program or may be composed of multiple computer programs. Furthermore, the answer generation program PG may be executed on a single computer or may be executed collaboratively by multiple computers. In addition, the answer generation program PG may partially use existing libraries.

[0033] The computer program, including the answer generation program PG, is provided on a non-temporary recording medium RM on which the computer program is recorded in a readable format. The recording medium RM is a portable memory such as a CD-ROM, USB memory, SD (Secure Digital) card, microSD card, or CompactFlash®. The control unit 101 reads various computer programs from the recording medium RM using a reading device (not shown in the figure) and stores the read computer programs in the storage unit 102. The computer programs stored in the storage unit 102 may also be provided via communication. In this case, the control unit 101 can acquire the computer programs via communication through the communication unit 103 and store the acquired computer programs in the storage unit 102.

[0034] The communication unit 103 is equipped with a communication interface for sending and receiving various types of data with external devices. A communication interface compliant with a communication standard such as LAN can be used as the communication interface of the communication unit 103. The external devices include a board processing device 120, a device data board 140, and a storage device 160 connected to the communication network NW1, as well as an LLM server 200 and a storage device 220 connected to the communication network NW2. When data to be transmitted is input from the control unit 101, the communication unit 103 transmits the data to the destination external device, and when data is received from the external device, it outputs the received data to the control unit 101.

[0035] The operation unit 104 is equipped with operating devices such as a touch panel, keyboard, and switches, and accepts various operations and settings from developers and others. The control unit 101 performs appropriate control based on the various operation information provided by the operation unit 104, and stores setting information in the storage unit 102 as needed.

[0036] The display unit 105 is equipped with a display device such as an LCD monitor or an OLED monitor, and displays information that should be notified to developers, etc., in response to instructions from the control unit 101.

[0037] The developer terminal 100 in this embodiment may be a single computer, or it may be a computer system composed of multiple computers and peripheral devices. Furthermore, the developer terminal 100 may be a virtualized virtual machine, or it may be a cloud.

[0038] The following describes an overview of the processes performed in the information processing system. Figures 4 to 9 are explanatory diagrams illustrating examples of response statements output to a developer's query. The developer terminal 100 receives the developer's query through a dedicated or general-purpose application program. The application program is, for example, the response generation program PG described above. Alternatively, a general-purpose web browser may be used. The developer terminal 100 can accept queries by text input through the operation unit 104. If the developer terminal 100 is equipped with a voice input unit, it may also accept queries by voice.

[0039] Figure 4 shows an example of a query received from a developer: "I want to know the wafer processing shape when using recipe Yyy with device XXX, so please tell me the combination of simulation models." In this example, "device XXX" represents the name of the substrate processing device 120, and "recipe Yyy" represents the type of recipe to be executed by "device XXX". Even if one searches a general website for the wafer processing shape when "recipe Yyy" is executed on "device XXX", it is highly unlikely that information specific to a particular device or recipe will be obtained. Similarly, even if one searches a general website for information in a highly specialized field such as simulation in substrate processing, it is highly unlikely that the information the developer desires will be obtained.

[0040] Therefore, in the information processing system according to this embodiment, a language model MD that references model data stored in the database 141 is used to generate a response statement to the developer's query. That is, when a developer's query is received, the developer terminal 100 generates input data (prompts) for the language model MD based on the received query, and sends the generated prompts to the LLM server 200 to generate a response statement.

[0041] For example, the developer terminal 100 generates a prompt specifying that its role is that of a process development specialist, that it should refer to model data stored in the database 141 as reference information, and that it should answer the developer's query. The developer terminal 100 may generate the prompt through internal processing only without obtaining modification and approval from the developer, or it may display the generated prompt on the display unit 105 and accept modification and approval from the developer.

[0042] The developer terminal 100 sends the generated prompt to the LLM server 200 via the communication networks NW1 and NW2. The LLM server 200 inputs the prompt received from the developer terminal 100 into the language model MD. The language model MD outputs a response statement according to the input prompt. In order to predict the state and quality of the substrate processing device, it is necessary to combine multiple simulation models hierarchically from the device information and recipe information. The appropriate combination of models differs depending on the device and process. In this embodiment, since the prompt specifies the reference of model data, the language model MD refers to the model data already collected in the device data base 140, selects an appropriate combination of models, and outputs the information of the selected combination of models as a response statement. The LLM server 200 returns the response statement data output from the language model MD to the developer terminal 100.

[0043] The developer terminal 100 receives the response data sent back from the LLM server 200 via the communication networks NW2 and NW1. Based on the received data, the developer terminal 100 displays the response on the display unit 105. For example, if the developer's query is a request for a combination of models to simulate the wafer processing shape when using recipe Yyy with apparatus XXX, the response will include a sentence indicating a combination of models suitable for realizing that simulation. The example in Figure 4 shows that a fluid model was selected as the model to simulate the gas phase reaction, a kinetic model as the model to simulate the transport phenomenon, and a stochastic shape model as the model to simulate the shape.

[0044] In the example of FIG. 4, an example of presenting a combination of models when an apparatus and a recipe are specified has been described. However, the conditions specified by the developer are not limited to the apparatus and the recipe, and may include the process type and the purpose of using the apparatus. For an internal developer, the process and the analysis purpose (such as the cause analysis of shape troubles or the elucidation of influences) are specified, and for an external developer, the process is specified as the purpose of using the apparatus.

[0045] FIG. 5 shows an example in which a developer's query "Please teach me the execution code because simulation is executed using the above combination of models" is received. This query represents a newly input developer's query after the response sentence in FIG. 4 is obtained (that is, a query newly input during the interaction between the developer and the LLM server 200). Based on the received query, the developer terminal 100 generates a prompt for the language model MD. In the prompt, a reference to the model data is specified as in the above. Also, the output format of the execution code may be specified in the prompt. The developer terminal 100 transmits the generated prompt to the LLM server 200.

[0046] The LLM server 200 inputs the prompt received from the developer terminal 100 to the language model MD. Since the LLM server 200 maintains the immediately preceding context while the session with the developer terminal 100 is maintained, when the above prompt is input to the language model MD, the language model MD can output a response sentence including the execution code for executing the simulation using the previously proposed combination of models.

[0047] The execution code for running the simulation varies in the modules (optional) used depending on the process type and purpose of use, and there are multiple types of physical equations and boundary conditions. Also, proper settings for numerical analysis such as the time interval of calculation and the interval of the mesh are required. Furthermore, setting parameters suitable for the device and process conditions is necessary. In the present embodiment, since the reference to the model data is specified in the prompt, the possibility of obtaining the execution code desired by the developer is increased by searching for the model data according to the device information, process type, and type of model used.

[0048] The LLM server 200 returns the data of the response sentence output from the language model MD to the developer terminal 100. Also, the LLM server 200 may create a file including the execution code for each model based on the response sentence output from the language model MD and store it in a storage unit (not shown) in the server as a file for downloading. Further, the LLM server 200 may transmit the created file to the developer terminal 100.

[0049] The developer terminal 100 receives the data of the response sentence returned from the LLM server 200 via the communication networks NW2 and NW1. The developer terminal 100 causes the display unit 105 to display the response sentence based on the received data. An example of the response sentence is as shown in FIG. 5. The response sentence includes, for example, a link for downloading a file of the execution code created by the language model MD. Also, when a simulation based on the execution code is executed in the developer terminal 100, the data regarding the execution result is transmitted to the device data base 140 and stored in the database 141 as model data.

[0050] Figures 6 and 7 show examples of interaction between a developer and the LLM server 200 when defining input information for a model. The example in Figure 6 shows a case where the developer's query to inquire about input information for the model is received after the definition of the input information for the model (variable name and numerical range) and conditions (plasma density 2 and plasma temperature 21 considering the ionization process) have been specified. Similar to the examples in Figures 4 and 5, the developer terminal 100 generates a prompt to be input to the language model MD based on the query received from the developer and sends the generated prompt to the LLM server 200. The prompt specifies a reference to model data.

[0051] The LLM server 200 inputs the prompt received from the developer terminal 100 into the language model MD. The language model MD outputs a response sentence referencing the model data. The LLM server 200 returns the response sentence data output from the language model MD to the developer terminal 100.

[0052] The developer terminal 100 receives the response data sent back from the LLM server 200 and displays the response on the display unit 105. The response shown in Figure 6 includes a statement indicating that the lonization variable can be set to 1, the plasma density (ne) to 2, and the plasma temperature (Te) to 21 as input information to the model. The developer can set the input conditions by referring to the response from the LLM server 200 and run the simulation.

[0053] The developer may enter a new query following the above response and request the LLM server 200 to provide the following response. Figure 7 shows an example where a query is entered requesting input information for apparatus A and process B when the ratio of plasma density to plasma temperature is constant and the pressure is halved, along with the input information set in the model.

[0054] The developer terminal 100 generates a prompt to be input to the language model MD based on the query received from the developer, and sends the generated prompt to the LLM server 200. The prompt specifies a reference to model data. The LLM server 200 inputs the prompt received from the developer terminal 100 into the language model MD. The language model MD outputs a response sentence that references the model data. The LLM server 200 returns the data of the response sentence output from the language model MD to the developer terminal 100.

[0055] The developer terminal 100 receives the response data sent back from the LLM server 200 and displays the response on the display unit 105. The response shown in Figure 7 includes a statement indicating that the lonization variable can be set to 1, the plasma density (ne) to 1, and the plasma temperature (Te) to 25.5 as new input information to the model. In this embodiment, since the language model MD is used, it is possible to set input information that takes into account the relationships between variables without having to input the relationship between pressure = density × temperature.

[0056] Figure 8 shows an example of receiving a developer query that reads, "I ran a simulation using the above execution code. The following data was obtained as an execution result. Please create an analysis script to analyze the execution result data." This query represents a newly entered developer query while maintaining the context of Figures 4 to 7, for example.

[0057] The developer terminal 100 generates input data (prompts) for the language model MD based on the received query. As described above, the model data is specified as reference information in the prompt. Furthermore, the prompt specifies a reference to the data obtained as a result of the simulation execution. The developer terminal 100 sends the generated prompts to the LLM server 200.

[0058] The LLM server 200 inputs the prompt received from the developer terminal 100 into the language model MD. As long as the session with the developer terminal 100 is maintained, the LLM server 200 maintains the context, so the language model MD outputs a response sentence that takes into account a series of pieces of information (e.g., device information, intended use, combination of models used for simulation). In the example in Figure 8, the creation of an analysis script is requested, so the response sentence is expected to include an analysis script. The LLM server 200 returns the data of the response sentence output from the language model MD to the developer terminal 100. Alternatively, the LLM server 200 may extract only the script portion from the response sentence output from the language model MD to create a script file and store it in the server's storage unit (not shown) as a downloadable file. Alternatively, the LLM server 200 may send the created script file to the developer terminal 100.

[0059] The developer terminal 100 receives the response data sent back from the LLM server 200 via the communication networks NW2 and NW1. Based on the received data, the developer terminal 100 displays the response on the display unit 105. An example of the response is shown in Figure 8. The response includes, for example, a link to download a script created by the LLM server 200.

[0060] Figure 9 shows an example of a developer query received: "Please create a report that includes a comparative explanation of the results of the substrate processing and the results of the simulation." This query represents a newly entered developer query while maintaining the context of Figures 4 to 8, for example.

[0061] The developer terminal 100 generates input data (prompts) for the language model MD based on the received query. As described above, the model data is specified as reference information in the prompt. Furthermore, the prompt specifies references to various data obtained by actually executing board processing in the board processing device 120 (for example, sensor data from various sensors). The developer terminal 100 sends the generated prompts to the LLM server 200.

[0062] The LLM server 200 inputs the prompt received from the developer terminal 100 into the language model MD. As long as the session with the developer terminal 100 is maintained, the LLM server 200 maintains the context, so the language model MD outputs a response sentence that takes into account a series of pieces of information (e.g., device information, intended use, etc.). In the example in Figure 9, a report including a comparative explanation of the results of the board processing and the results of the simulation is requested, so the output of a report including a sentence that compares the results of the board processing and the results of the simulation is expected. The LLM server 200 sends the response sentence data output from the language model MD back to the developer terminal 100.

[0063] The developer terminal 100 receives the response data sent back from the LLM server 200 via the communication networks NW2 and NW1. Based on the received data, the developer terminal 100 displays the response on the display unit 105. An example of the response is shown in Figure 9. The response (report) includes a comparative explanation of the results of the board processing and the results of the simulation.

[0064] Figures 4 to 9 show the developer's queries and the LLM server 200's responses that occurred during a series of interactions between the developer and the LLM server 200. However, it is not essential to utilize the interaction history. In other words, the LLM server 200 may generate a response to the developer's query without referring to the interaction history and send the generated response data to the developer terminal 100.

[0065] Figure 10 is a flowchart illustrating the procedure of processing performed by the developer terminal 100. When a developer asks a question regarding process development, they launch a dedicated application (e.g., answer generation program PG) or a web browser on the developer terminal 100 and access the LLM server 200. The developer terminal 100 receives the developer's query through the operation unit 104 (step S101). The control unit 101 of the developer terminal 100 generates input data (prompts) to be input to the language model MD based on the query received in step S101 (step S102).

[0066] At this time, the control unit 101 generates a prompt specifying that the role is that of a process development expert, that it should refer to model data stored in the database 141 of the device data base 140 as reference information, and that it should answer the developer's query. The prompt may also be generated using template data. For example, if template data is stored in the storage unit 102 and a developer's query is received, the control unit 101 can generate a prompt by embedding the developer's query into the template data read from the storage unit 102.

[0067] The control unit 101 transmits the generated prompt to the LLM server 200 via the communication unit 103 (step S103). The transmitted prompt reaches the LLM server 200 via the communication networks NW1 and NW2.

[0068] Upon receiving a prompt, the LLM server 200 inputs the received prompt into the language model MD. The language model MD outputs a response sentence according to the input prompt. At this time, model data stored in the database 141 of the device data infrastructure 140 is referenced as needed. The LLM server 200 sends the response sentence data output from the language model MD back to the developer terminal 100.

[0069] The control unit 101 of the developer terminal 100 receives the response data sent back from the LLM server 200 via the communication unit 103 (step S104). Based on the received response data, the control unit 101 displays the response on the display unit 105 (step S105). If the query received from the developer is, for example, a request for a model combination, a simulation execution code, input information for the model, an analysis script, or a report, then a response text like those shown in Figures 4 to 9 will be obtained, and the corresponding response text will be displayed on the display unit 105 of the developer terminal 100. The same applies even if the query received from the developer is not a request for creation as described above, and a response text corresponding to the developer's query will be displayed on the display unit 105 of the developer terminal 100.

[0070] Alternatively, the developer terminal 100 may sequentially send multiple prompts to the LLM server 200 and receive a response message for each prompt from the LLM server 200. For example, if the control unit 101 receives a combination of models and a simulation execution code as a response message, it can execute a simulation that couples multiple models based on the received combination of models and execution code. For example, if a fluid model and a stochastic shape model are selected as the combination of models, and the execution code for each model is obtained as a response message, the control unit 101 will execute a simulation using the fluid model to calculate the state of the chamber, and then use the calculation results to calculate the processed shape of the substrate using the stochastic shape model. The control unit 101 can display the state of the chamber using the fluid model and the substrate shape calculated by coupling the fluid model and the stochastic shape model on the display unit 105.

[0071] Furthermore, when the control unit 101 receives a model combination, simulation execution code, and analysis script as a response, it can automate everything from the execution of the simulation to a comparative explanation with the actual experiment. If a discrepancy occurs between the simulation execution result and the actual experiment, the control unit 101 may display a text on the display unit 105 as a response explaining the cause of the equipment malfunction, etc.

[0072] As described above, in this embodiment, a response statement based on the results of simulations performed in the past (model data), according to the device information and / or intended use of the substrate processing apparatus 120, can be generated using the language model MD and presented to the developer.

[0073] The matters described in each embodiment can be combined with each other. Furthermore, the independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. The claims may be described in a form in which claims refer to two or more other claims (multi-claim form), or in a form in which multi-claims refer to at least one multi-claim (multi-multi-claim).

[0074] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the invention is indicated by the claims, not in the sense described above, and all modifications within the sense and scope equivalent to the claims are intended.

[0075] For example, in this embodiment, a developer terminal 100 connected to a communication network NW1 receives a question from a developer, and the developer terminal 100 generates an answer using the LLM server 200. However, a server that generates an answer using the LLM server 200 may be provided separately within the communication network NW1. In this case, the developer's question received by the developer terminal 100 is sent to the server within the communication network NW1, the server generates an answer using the LLM server 200, and the generated answer is sent back to the developer terminal 100.

[0076] Furthermore, the above server may also accept questions from a general user's terminal device (not shown) connected to the communication network NW2. In this case, the general user's terminal device sends the question to the server via the communication networks NW2 and NW1, and the server that receives the question generates an answer using the LLM server 200 and sends the generated answer back to the general user's terminal device via the communication networks NW1 and NW2.

[0077] 100 Developer terminal 101 Control unit 102 Storage unit 103 Communication unit 104 Operation unit 105 Display unit 200 LLM server PG Answer generation program RM Recording medium MD Language model

Claims

1. An information processing method that, using a computer, receives device information and / or intended use of a substrate processing device, generates a response corresponding to the received device information and / or intended use using a language model that references output data obtained by performing a simulation related to substrate processing, and outputs the generated response.

2. The information processing method according to claim 1, wherein the computer performs a process to generate a response statement that includes selection information for one or more types of simulation models according to the device information and / or the purpose of use.

3. The information processing method according to claim 2, wherein the computer performs a process to generate a response statement including executable code for performing a simulation using one or more types of simulation models, according to the device information and / or the intended use.

4. The information processing method according to claim 2, wherein the computer performs a process to generate a response statement that includes an analysis script for analyzing the results of a simulation performed by one or more types of simulation models, according to the device information and / or the purpose of use.

5. The information processing method according to any one of claims 2 to 4, wherein the computer receives a query from a user requesting the aforementioned response, and based on the received query, the computer performs a process to generate input data to be input to the language model.

6. The information processing method according to claim 5, wherein the input data includes one of the following, which is designated as reference information for the language model: model data, simulation execution results, and substrate processing execution results.

7. The information processing method according to any one of claims 1 to 4, wherein the computer performs a process to obtain the execution result of substrate processing by the substrate processing apparatus and to generate a report as the response statement, which includes a comparative explanation between the obtained execution result and the execution result of a simulation that simulates the substrate processing.

8. The information processing method according to claim 1, wherein the computer performs a process of coupling and executing a first simulation using a first simulation model and a second simulation using a second simulation model different from the first simulation model, and generating an explanatory text using the language model that includes the degree of influence of the first simulation on the execution result of the second simulation.

9. An information processing device comprising one or more processors, wherein the processor receives device information and / or purpose of use of a substrate processing device, generates a response statement corresponding to the received device information and / or purpose of use using a language model that refers to a group of output data obtained by executing a simulation related to substrate processing, and outputs the generated response statement.

10. A computer program that receives device information and / or intended use of a substrate processing device, generates a response corresponding to the received device information and / or intended use using a language model that references output data obtained by running a simulation related to substrate processing, and causes a computer to execute a process to output the generated response.

Images