Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 11/09/2023, 07/08/2024, and 09/30/2025 were filed before the mailing date of the first office action. The submissions are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101. Claims 1-10 are directed to a system, claims 11-16 are directed to a method, and claims 17-20 are directed to a non-transitory computer readable media; therefore, claims 1-20 fall within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter). However, claims 1-20 fall within the judicial exception of an abstract idea, specifically the abstract ideas of “Mental Processes” (including observation, evaluation, and opinion) and “Mathematical Concepts (including mathematical calculations and relationships)”.
Claim 1:
Claim 1 is directed to a system; therefore, the claim does fall within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Claim 1 recites the following abstract ideas:
processing the plurality of input characters to determine an intent of the user prompt request (mental step directed to observation, evaluation – a person could process a plurality of observed input characters to determine an intent of an observed user prompt request);
generating a refined prompt based on performing a mark-up language transform on the plurality of input characters and the intent (mental step directed to observation, evaluation – a person could generate a refined prompt in their mind by marking up, or transforming an observed input according to an observed or determined prompt intent, potentially assisted by pen and paper (see MPEP 2106.04(a)(2)(III))).
Claim 1 recites the following additional elements:
providing a user interface to a user computing system, wherein the user interface comprises an integrated development environment; obtaining a plurality of input characters from the user computing system via the user interface, wherein the plurality of input characters are descriptive of a user prompt request; providing the refined prompt to a generative model to receive a generative output. Providing a user interface, obtaining a plurality of input characters, and providing a refined prompt to a generative model are all interpreted as transmitting and receiving data over a network in the field of user or technological environment in which the claimed abstract ideas are performed. The user interface and integrated development environment are interpreted as generic computer components used to merely implement the claimed abstract ideas. These additional elements do not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(d)(II), MPEP 2106.05(f) and MPEP 2106.05(h)).
Claim 11 is a method claim and its limitation is included in claim 1. The only difference is that claim 11 requires a method of implementing the abstract ideas identified in the analysis of claim 1. Therefore, claim 11 is rejected for the same reasons as claim 1.
Claim 17 is a non-transitory computer-readable media claim and its limitation is included in claim 1. The only difference is that claim 17 requires a non-transitory computer-readable media; which is interpreted as a generic computer component used to merely implement the abstract ideas identified in the analysis of claim 1 (see MPEP 2106.05(f)). Therefore, claim 17 is rejected for the same reasons as claim 1.
The independent claims are not patent eligible.
Dependent claims 2-10, 12-16, and 18-20 when analyzed as a whole are held to be patent ineligible under 35 U.S.C. 101 because the additional recited limitations fail to establish that the claims are not directed to an abstract idea, as they recite further embellishment of the judicial exception.
Claim 2 recites wherein the operations further comprise: receiving the generative output from the generative model; and providing the generative output to the user computing system. Receiving an output and providing an output to a user system are both interpreted as an additional element directed to receiving and transmitting data over a network in the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(d)(II) and MPEP 2106.05(h)).
Claim 3 recites wherein the operations further comprise: processing the plurality of input characters to determine a plurality of text tokens associated with a plurality of input character sets determined to be semantically linked; and providing a plurality of respective token indicators associated with at least a subset of the plurality of text tokens, wherein each respective token indicator comprises a graphical indicator indicating a length and location of a respective text token. Determining a plurality of text tokens and determining input character sets that are semantically linked are interpreted as mental steps directed to observation, evaluation – a person could determine which observed input character sets are semantically linked in their mind and determine text tokens for the observed character sets in their mind. Providing token indicators comprising graphical indicators is interpreted as transmitting data over a network in the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(d)(II) and MPEP 2106.05(h)).
Claim 4 recites wherein the integrated development environment is configured to receive the plurality of input characters and is configured to perform the mark-up language transform. Performing the mark-up language transform is interpreted as a mental step directed to observation, evaluation – a person could mark up, or transform an observed input, potentially assisted by pen and paper. Receiving a plurality of input characters is interpreted as an additional element directed to receiving data over a network in the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(d)(II) and MPEP 2106.05(h)).
Claim 5 recites wherein the integrated development environment is associated with prompt-generation mark-up language. This limitation is interpreted as a further description of the kind of generic computer component used to apply the claimed abstract ideas and the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(f) and MPEP 2106.05(h)).
Claim 6 recites wherein the prompt-generation mark-up language comprises one or more delimiters selected based on a determined low likelihood of use in traditional natural language. This limitation is interpreted as a further description of the kind of mark-up language a person could use to transform an observed user prompt in their mind, as a person could mentally utilize delimiters to determine words that have a low likelihood of use, potentially assisted by pen and paper.
Claim 7 recites wherein the integrated development environment is associated with a text-encoding system associated with a set of pre-determined symbols associated with a set of formatting operators. This limitation is interpreted as a further description of the kind of generic computer component used to apply the claimed abstract ideas and the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(f) and MPEP 2106.05(h)).
Claim 8 recites wherein the refined prompt comprises a preamble associated with a specified task. Wherein the refined prompt comprises a preamble associated with a task is interpreted as a further description of the kind of refined prompt that can be mentally generated.
Claim 9 recites wherein the refined prompt comprises a body associated with one or more details to include in the generative output. Wherein the refined prompt comprises a body associated with more detail is interpreted as a further description of the kind of refined prompt that can be mentally generated.
Claim 10 recites wherein the operations further comprise: determining one or more prompt term suggestions based on the intent; and providing the one or more prompt term suggestions as selectable user interface elements. Determining one or more prompt term suggestions based on an intent is interpreted as a mental step directed to observation, evaluation – a person could determine prompt term suggestions in their mind based on an observed or determined intent. Providing the prompt term suggestions as selected user interface elements is interpreted as an additional element directed to displaying, or transmitting information over a network in the technological environment in which the abstract ideas are performed, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea (see MPEP 2106.05(d)(II) and MPEP 2106.05(h)).
Claim 12 recites wherein the one or more prompt term suggestions are determined based on a determined intent of the prompt request, wherein the determined intent is determined based on processing at least a subset of the plurality of input characters. Determining prompt suggestions based on a determined intent and processing the input characters is interpreted as a mental step directed to observation, evaluation – a person could determine prompt suggestions and a prompt intent in their mind based on an observed plurality of input characters.
Claim 13 recites wherein the one or more prompt term suggestions are obtained from an index of prompt terms. Determining prompt suggestions based on an index of prompt terms is interpreted as a mental step directed to observation, evaluation – a person could determine prompt suggestions in their mind based on an observed or mentally determined index of prompt terms.
Claim 14 recites wherein the index of prompt terms was generated based on historical prompt data associated with historical content generation. Generating an index of prompt terms based on historical prompt data is interpreted as a mental step directed to observation, evaluation – a person could generate an index of prompt terms in their mind based on an observed or mentally determined historical prompt data.
Claim 15 recites wherein the index of prompt terms was generated based on one or more training labels associated with the training dataset for the generative model. Generating an index of prompt terms based on historical prompt data is interpreted as a mental step directed to observation, evaluation – a person could generate an index of prompt terms in their mind based on an observed or mentally determined labels associated with training data.
Claim 16 recites wherein the plurality of input characters comprise a first structure, and wherein the refined prompt comprises a second structure. Wherein the input characters comprise a first structure and the refined prompt comprises a second structure is interpreted as a further description of the kind of input characters that can be observed and the kind of refined prompt that can be mentally generated.
Claim 18 recites wherein the plurality of input characters are descriptive of a subject and one or more details to include in a generated subject, wherein the refined prompt comprises a restructured text string descriptive of a predetermined style, and wherein the refined prompt is descriptive of the subject and the one or more details. Wherein the input characters are descriptive of a subject and included details and the refined prompt comprises a descriptive restructured text string with a predetermined style is interpreted as a further description of the kind of input characters that can be observed and the kind of refined prompt that can be mentally generated.
Claim 19 recites wherein generating the refined prompt comprises word mapping, wherein a subset of the plurality of input characters are mapped to one or more alternate words. Generating a refined prompt by mapping input characters to one or more alternate words is interpreted as a mental step directed to observation, evaluation – a person could generate a refined prompt in their mind by mapping input characters to one or more alternate words in their mind, potentially assisted by pen and paper.
Claim 20 recites wherein generating the refined prompt comprises structure mapping, wherein a subset of the plurality of input characters are mapped to a predefined structure associated with a preamble and a body of the refined prompt. Generating a refined prompt by mapping input characters to a predefined structure is interpreted as a mental step directed to observation, evaluation – a person could generate a refined prompt in their mind by mapping input characters to a predefined structure comprising a preamble and a body in their mind, potentially assisted by pen and paper.
Viewed as a whole, these additional claim elements do not provide meaningful limitations to transform the abstract idea into a patent eligible application of the abstract idea such that the claims amount to significantly more than the abstract idea itself. Therefore, the claims are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 4-5, and 8-20 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al* (US 20220229832 A1, herein Li) in view of Dang et al* (“How to Prompt? Opportunities and Challenges of Zero- and Few-Shot Learning for Human-AI Interaction in Creative Applications of Generative Models”, herein Dang).
*this document was included in the IDS dated 07/08/2024
Regarding claim 1, Li teaches a computing system, the system comprising: one or more processors; and one or more non-transitory computer-readable media that collectively store instructions that, when executed by the one or more processors (para. [0046] recites “Storage system 1005 may comprise any computer readable storage media readable by processing system 1020 and capable of storing software 1010. Storage system 1005 may include volatile and nonvolatile, removable and nonremovable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, cache memory or other data”), cause the computing system to perform operations, the operations comprising:
providing a user interface to a user computing system, wherein the user interface comprises an integrated development environment (para. [0016] recites “The disclosed system may be implemented with any content development application. The solution includes a complete, natural language generation modelling powered solution to allow users to generate content with minimal text inputs in an iterative fashion”. Para. [0019] recites “the user system design components 135 may be cloud based and access using a user interface on user system 105” (i.e., a user interface integrated with a content development application, or environment));
obtaining a plurality of input characters from the user computing system via the user interface, wherein the plurality of input characters are descriptive of a user prompt request; processing the plurality of input characters to determine an intent of the user prompt request (para. [0020] recites “The query understanding component 140 is used to process the user query and make determinations about the user's request. The query understanding component 140 takes the text query input by the user (i.e., the user query) and tries to understand the user's intention” (i.e., processing an obtained user input prompt comprising text, or a plurality of characters, and determining the intent of the user prompt));
generating a refined prompt based on [performing a mark-up language transform on] the plurality of input characters and the intent (para. [0030] recites “The prompt design component 115 may be used to generate an appropriate and ideally the best prompt to the natural language generation model 125 such that the desired output is generated by the model” (i.e., improving, or refining the user prompt)); and
providing the refined prompt to a generative model to receive a generative output (para. [0033] recites “At step 320, the prompt is provided to a natural language generation model (e.g., natural language generation model 125), such as GPT-3. The natural language generation model performs modelling, and at step 325 the output from the model is received. At step 330, the output is used to generate response content in a format compatible with the content generation application” (i.e., providing the improved, or refined prompt to a generative model and receiving an output from the model)).
However, while Li teaches that mark-up language can be used in at least paragraph [0056], Li does not explicitly teach performing a mark-up language transform on an input prompt.
Dang teaches performing a mark-up language transform on an input prompt (the description of fig. 6 recites “This GUI example treats the presentation of prompts similar to markdown, LaTeX or code editors: The GUI is split between prompt view and text view, allowing users to switch between editorial and evaluative tasks” (i.e., applying a mark-up language transform on a prompt)).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine these teachings by applying the mark-up language transforms from Dang to refine and improve the input prompts from Li. Li and Dang are both directed to systems which edit user input prompts to a generative model. One of ordinary skill in the art would be motivate to utilize the mark-up language transformation capability from Dang to provide an additional option for editing and improving the prompts from Li.
Regarding claim 2, the combination of Li and Dang teaches the system of claim 1, wherein the operations further comprise: receiving the generative output from the generative model; and providing the generative output to the user computing system (Li para. [0033] recites “At step 320, the prompt is provided to a natural language generation model (e.g., natural language generation model 125), such as GPT-3. The natural language generation model performs modelling, and at step 325 the output from the model is received. At step 330, the output is used to generate response content in a format compatible with the content generation application. The response content may be suggested content that are provided in one or more options to the user for selection by the user. The response content may be displayed at step 335” (i.e., receiving an output from the model and providing it to the user)).
Regarding claim 4, the combination of Li and Dang teaches the system of claim 1, wherein the integrated development environment is configured to receive the plurality of input characters and is configured to perform the mark-up language transform (Li para. [0016] recites “The disclosed system may be implemented with any content development application. The solution includes a complete, natural language generation modelling powered solution to allow users to generate content with minimal text inputs in an iterative fashion”. Li para. [0020] recites “The query understanding component 140 is used to process the user query and make determinations about the user's request. The query understanding component 140 takes the text query input by the user (i.e., the user query) and tries to understand the user's intention”. The description of fig. 6 of Dang recites “This GUI example treats the presentation of prompts similar to markdown, LaTeX or code editors: The GUI is split between prompt view and text view, allowing users to switch between editorial and evaluative tasks” (i.e., applying a mark-up language transform on a received prompt by the development application, or environment)).
Regarding claim 5, the combination of Li and Dang teaches the system of claim 1, wherein the integrated development environment is associated with prompt-generation mark-up language (the description of fig. 6 of Dang recites “This GUI example treats the presentation of prompts similar to markdown, LaTeX or code editors: The GUI is split between prompt view and text view, allowing users to switch between editorial and evaluative tasks”. Li para. [0016] recites “The disclosed system may be implemented with any content development application. The solution includes a complete, natural language generation modelling powered solution to allow users to generate content with minimal text inputs in an iterative fashion” (i.e., utilizing mark-up language on prompts in a content development application, or environment)).
Regarding claim 8, the combination of Li and Dang teaches the system of claim 1, wherein the refined prompt comprises a preamble associated with a specified task (Dang fig. 1 depicts a refined prompt with a task-orient preamble associated with translation)).
Regarding claim 9, the combination of Li and Dang teaches the system of claim 1, wherein the refined prompt comprises a body associated with one or more details to include in the generative output (Dang fig. 1 depicts a refined prompt comprising a body associated with details of the tone for the translation task output)).
Regarding claim 10, the combination of Li and Dang teaches the system of claim 1, wherein the operations further comprise: determining one or more prompt term suggestions based on the intent; and providing the one or more prompt term suggestions as selectable user interface elements (Li para. [0027] recites “a design or other tool from the content generation application 130 may be used to provide options or new designs or layout at step 216. As an example, the user query may be "make the background purple." This may result in the design tools offering several shades of purple background options. The user may select whether or which options to keep at decision block 218”. Li para. [0030] recites “The prompt design component 115 may use initial prompt design 250 and good prompt examples 228 as input to a prompt library 230 to generate possible prompts” (i.e., providing prompt suggestions using the user interface)).
Claim 11 is a method claim and its limitation is included in claim 1. The only difference is that claim 11 requires a method (Li para. [0002] recites “One general aspect includes a computer-implemented method for automatically generating intelligent content”). Therefore, claim 11 is rejected for the same reasons as claim 1.
Regarding claim 12, the combination of Li and Dang teaches the method of claim 11, wherein the one or more prompt term suggestions are determined based on a determined intent of the prompt request, wherein the determined intent is determined based on processing at least a subset of the plurality of input characters (Li para. [0030] recites “The query understanding component 140 is used to process the user query and make determinations about the user's request. The query understanding component 140 takes the text query input by the user (i.e., the user query) and tries to understand the user's intention”. Li para. [0030] recites “The prompt design component 115 may use initial prompt design 250 and good prompt examples 228 as input to a prompt library 230 to generate possible prompts” (i.e., determining a prompt intent based on the text, or characters of the prompt and determining prompt suggestions based on the determined intent)).
Regarding claim 13, the combination of Li and Dang teaches the method of claim 11, wherein the one or more prompt term suggestions are obtained from an index of prompt terms (Li para. [0020] recites “The query understanding component 140 is used to process the user query and make determinations about the user's request. The query understanding component 140 takes the text query input by the user (i.e., the user query) and tries to understand the user's intention”. Li para. [0023] recites “Once the user system design components 135 have processed the user query and determined what action the user is requesting, the action or user query is sent to the application/service component 110, which may be cloud based. The application/service component 110 may send the user query or action to the prompt design component 115. The prompt design component 115 is used to generate a prompt that is appropriate for input to the natural language generation model 125. The prompt design component 115 may be an AI component that uses a machine learning algorithm or neural network to develop better prompts over time. The prompt design component 115 may access the knowledge repositories 120 including user preference data, a prompt library, and prompt examples to generate the prompt and return it to the application/service component 110” (i.e., determining prompt suggestions based on an index, or library of prompt data)).
Regarding claim 14, the combination of Li and Dang teaches the method of claim 13, wherein the index of prompt terms was generated based on historical prompt data associated with historical content generation (Li para. [0020] recites “The query understanding component 140 is used to process the user query and make determinations about the user's request. The query understanding component 140 takes the text query input by the user (i.e., the user query) and tries to understand the user's intention”. Li para. [0027] recites “Intent detection may include determining the intent of the user query. The intent detection step 212 may include using the user preference history, current deck global information, and/or current deck edit history from data 234 to determine the intent” (i.e., determining prompt suggestions based on a library of previous, or historical prompt data)).
Regarding claim 15, the combination of Li and Dang teaches the method of claim 13, wherein the index of prompt terms was generated based on one or more training labels associated with the training dataset for the generative model (Li para. [0023] recites “The prompt design component 115 may be an AI component that uses a machine learning algorithm or neural network to develop better prompts over time. The prompt design component 115 may access the knowledge repositories 120 including user preference data, a prompt library, and prompt examples to generate the prompt and return it to the application/service component 110” Li para. [0030] recites “Many prompts may also be learned from the user keeping suggestions (e.g., continuous learning). The possible prompts may be ranked by a prompt ranker based at least in part on user preference history, current deck global information, and/or current deck edit history from data 234 to generate a ranked list of prompts” (i.e., determining prompt terms based on labeled past examples, or training data, that the model can use and reuse)).
Regarding claim 16, the combination of Li and Dang teaches the method of claim 11, wherein the plurality of input characters comprise a first structure, and wherein the refined prompt comprises a second structure (Dang figure 1 depicts wherein the input prompt has a first structure and the refined prompt has a second structure)).
Claim 17 is a non-transitory computer-readable media claim and its limitation is included in claim 1. The only difference is that claim 17 requires a non-transitory computer-readable media (Li para. [0046] recites “Storage system 1005 may comprise any computer readable storage media readable by processing system 1020 and capable of storing software 1010. Storage system 1005 may include volatile and nonvolatile, removable and nonremovable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, cache memory or other data”). Therefore, claim 17 is rejected for the same reasons as claim 1.
Regarding claim 18, the combination of Li and Dang teaches the one or more non-transitory computer-readable media of claim 17, wherein the plurality of input characters are descriptive of a subject and one or more details to include in a generated subject, wherein the refined prompt comprises a restructured text string descriptive of a predetermined style, and wherein the refined prompt is descriptive of the subject and the one or more details (Dang fig. 1 shows wherein an input prompt has a subject and required details associated with the requested translation task, and a restructured prompt with a predetermined tone, or style, and details associated with the translation task, or subject description)).
Regarding claim 19, the combination of Li and Dang teaches the one or more non-transitory computer-readable media of claim 17, wherein generating the refined prompt comprises word mapping, wherein a subset of the plurality of input characters are mapped to one or more alternate words (Dang fig. 1 shows wherein an input prompt can be mapped to alternate words, for example, the German translation task includes mappings to French, Spanish, and Greek alternatives)).
Regarding claim 20, the combination of Li and Dang teaches the one or more non-transitory computer-readable media of claim 17, wherein generating the refined prompt comprises structure mapping, wherein a subset of the plurality of input characters are mapped to a predefined structure associated with a preamble and a body of the refined prompt (Dang fig. 1 shows wherein the refined prompt has mapped the input prompt to a predefined structure including the “translate” preamble and body comprising details of the requested translation task)).
Claims 3 and 7 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al* (US 20220229832 A1, herein Li) in view of Dang et al* (“How to Prompt? Opportunities and Challenges of Zero- and Few-Shot Learning for Human-AI Interaction in Creative Applications of Generative Models”, herein Dang), in further view of Heller et al (US 20240242037 A1, herein Heller).
Regarding claim 3, the combination of Li and Dang teaches the system of claim 1.
However, the combination of Li and Dang does not teach wherein the operations further comprise: processing the plurality of input characters to determine a plurality of text tokens associated with a plurality of input character sets determined to be semantically linked; and providing a plurality of respective token indicators associated with at least a subset of the plurality of text tokens, wherein each respective token indicator comprises a graphical indicator indicating a length and location of a respective text token.
Heller teaches wherein the operations further comprise: processing the plurality of input characters to determine a plurality of text tokens associated with a plurality of input character sets determined to be semantically linked; and providing a plurality of respective token indicators associated with at least a subset of the plurality of text tokens, wherein each respective token indicator comprises a graphical indicator indicating a length and location of a respective text token (para. [0023] recites “techniques and mechanisms described herein provide for the division of text into chunks, and the incorporation of those chunks into prompts that can be provided to a large language model. For instance, a large language model may impose a limit of, for instance, 8,193 tokens on a task, including text input, text output, and task instructions”. Para. [0101] recites “In the event that the text portion fits into the last text chunk size, the text portion is inserted into the last text chunk at 610. If instead the text portion is the first to be processed, or the text portion does not fit into the last text chunk size, then the text portion is inserted into a new text chunk at 612” (i.e., determining a plurality of text tokens that have been determined to be semantically linked and wherein a token has an associated indicator for length and location)).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine these teachings by utilizing the text division technique from Heller to the input prompt editing system from Li (as modified by Dang). Li and Heller are both directed to systems which can edit and refine user prompts to a generative model. One of ordinary skill in the art would be motivate to utilize Heller’s text parsing and token generation method to section long user prompts from Li’s prompt editing system into sections with more manageable lengths, as at least paragraph [0029] of Heller recites “text chunking may reduce token overhead and hence cost expended on large language model prompts”.
Regarding claim 7, the combination of Li and Dang teaches the system of claim 1, and the integrated development environment (see at least paragraph [0016] of Li).
However, the combination of Li and Dang does not explicitly teach a text-encoding system associated with a set of pre-determined symbols associated with a set of formatting operators.
Heller teaches a text-encoding system associated with a set of pre-determined symbols associated with a set of formatting operators (para. [0036] recites “Parsed input text is determined based on the original input text at 104. In some embodiments, determining the original input text may involve performing one or more text processing operations such as cleaning, sharding, chunking, and the like. Additional details regarding text processing operations are discussed throughout the application as filed, for instance with respect to FIG. 2, FIG. 3, FIG. 4, FIG. 5, and FIG. 6”. Para. [0119] recites “parsing the chat message at 818 may involve searching the chat response message 816 for the natural language text and/ or the one or more skill codes. Skill codes identified in this way may be used to influence the generation of the chat output message sent at 822” (i.e., encoding text with predetermined formatting codes, or operators)).
See claim 3 for motivation to combine.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Li et al* (US 20220229832 A1, herein Li) in view of Dang et al* (“How to Prompt? Opportunities and Challenges of Zero- and Few-Shot Learning for Human-AI Interaction in Creative Applications of Generative Models”, herein Dang), in further view of Nakazawa et al (US 20110202532 A1, herein Nakazawa).
Regarding claim 6, the combination of Li and Dang teaches the system of claim 5, including the prompt-generation mark-up language (see at least figure 6 of Dang).
However, the combination of Li and Dang does not explicitly teach delimiters selected based on a determined low likelihood of use in traditional natural language.
Nakazawa teaches delimiters selected based on a determined low likelihood of use in traditional natural language (para. [0038] recites “The word dictionary storage unit 221 is a dictionary for words used by the specified section text analysis means 13 to perform a linguistic analysis to the specified section text. In the dictionary of each word, the information necessary for the linguistic analysis process in the specified section linguistic analysis means 13 is stored among dictionary information used in general natural language processing techniques, such as grammar information including notation of a word, word delimiter, word class and conjugation, grammar information indicating a method of connection between words, statistical information, information of a stop word indicating whether each word is important for usage and purpose upon carrying out the present invention, and dictionary information used in general natural language processing techniques” (i.e., delimiters indicating a level of importance, or likelihood of usage, for given words in a natural language setting)).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine these teachings by utilizing the delimiters from Nakazawa to modify the mark-up language from Dang to determine the likelihood of suggested words in the prompt refining process from Li. Li and Nakazawa both teach referencing a prompt, or word library for natural language processing task. One of ordinary skill would be motivated to mark when the system from Li suggested a potential prompt with unnatural or archaic language using a word dictionary such as the one taught by Nakazawa.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
“LINGUIST: Language Model Instruction Tuning to Generate Annotated Utterances for Intent Classification and Slot Tagging” (Rosenbaum et al) teaches a method for generating annotated data for prompt intent classification and slot tagging data generation.
“PromptGen: Automatically Generate Prompts using Generative Models” (Zhang et al) teaches a method for automatically generate prompts based on inputs by leveraging a pre-trained generative model.
“Induced Natural Language Rationales and Interleaved Markup Tokens Enable Extrapolation in Large Language Models” (Bueno et al) teaches inducing explanations and markup tokens using in-context learning for a generative transformer model.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to LEAH M FEITL whose telephone number is (571) 272-8350. The examiner can normally be reached on M-F 0900-1700 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached on (571) 270-5871. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/L.M.F./ Examiner, Art Unit 2147
/VIKER A LAMARDO/Supervisory Patent Examiner, Art Unit 2147
Prosecution Insights