DETAILED ACTION
This office action is in response to Applicant’s Amendment/Request for Reconsideration, received on 03/03/2026. Claims 1-14, 18-20 have been amended. Claims 1-20 are pending and have been considered.
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 .
Response to Arguments
Applicant’s arguments, see pg. 14, filed 03/03/2026, with respect to “Drawings” have been fully considered and are persuasive. The objections of the drawings has been withdrawn.
Applicant’s arguments, see pgs. 14-15, filed 03/03/2026, with respect to “Specification” have been fully considered and are persuasive. The objection of the specification has been withdrawn.
Applicant’s arguments, see pg. 15, filed 03/03/2026, with respect to “Claim Objections” have been fully considered and are persuasive. The objections of claims 1-9 have been withdrawn.
Applicant’s arguments, see pg. 15, filed 03/03/2026, with respect to “Claim Rejections Under 35 U.S.C. 112” have been fully considered and are persuasive. The rejections of claims 2-3, 11-12, and 20 under 35 U.S.C. 112(b) have been withdrawn.
Applicant's arguments filed 03/03/2026, see pgs. 15-17, with respect to “Rejections Under 35 U.S.C. 101” have been fully considered but they are not persuasive.
Applicant’s representative asserts, “Without acquiescing to the rejections, and solely in the interest of advancing prosecution, Applicant has amended the claims. Accordingly, at least as amended, the claims are not directed to any abstract ideas, incorporate any alleged abstract idea into a practical application thereof, and include additional features amounting to significantly more than any alleged abstract idea.
For example, as discussed in greater detail throughout the application, the features of the pending claims provide a ‘technical solution relat[ing] to perspective-based validation of prompts to generative artificial intelligence (Al).’ For example, users may ‘expect generative Al to provide detailed answers in an increasingly broad range of domains...that are related to sensitive or confidential data.’ Accordingly, ‘[t]his technical solution can identify and filter sensitive information from a user query to generate a prompt for a generative Al.’ (See Application, paragraph [0018]). Some implementations of ‘[t]he technical solution...can also generate a validated prompt for a given query, or class of queries, that has a lower risk of containing false information,’ and/or ‘modify or expand a prompt based on the query to include additional information.’ (See Application, paragraphs [0019]-[0020]). Further, some implementations of the technical solution may provide one or more prompts according to an original query to one or more generative Al systems, and determine a divergence between responses.’ Beneficially, at least these features of the technical solution may ‘determine that [] diverging results are likely to be a “hallucination” of generative Al, and can discard or downrank those results’ and may ‘validate many prompts with respect to many queries, and can store validated prompts for later usage.’ (See Application, paragraph [0021]).
Thus, the features of the pending claims may provide several technical improvements in the field of generative Al. For example, the claimed technical solution may ‘provide a technical improvement to increase accuracy and decrease output corresponding to “Al hallucination” in response to increasingly complex and targeted queries.’ (See Application, paragraph [0021]). Additionally or alternatively, ‘the system can execute a plurality of differently tuned or trained large language models each to generate the responses, to provide a technical improvement to test queries for convergence or divergence across a plurality of large language models. Thus, the technical solution can provide computational efficiency in identifying responses with veracity, by imputing veracity to convergent responses that can be generated at a rate and at a level of verifiability beyond the capability of manual processes.’ (See Application, paragraph [0022]).
Amended independent claim 1 reflects this technical solution by reciting, in part, ‘receiv[ing], via a user interface, a first prompt for a large language model including a first query that references first data’, ‘generat[ing] a plurality of second prompts for the large language model based on the first prompt and the first data, each of the plurality of second prompts including one or more second data clarifying the first query’, ‘generat[ing], by the large language model receiving the plurality of second prompts, one or more respective responses to each of the plurality of second prompts’, ‘validat[ing] one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompts meets an accuracy threshold’, ‘caus[ing] the user interface to present the one or more validated second prompts or one or more responses to the one or more validated second prompts,’ and ‘select[ing] a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input.’
At least these features of claim 1, individually or in combination, provide at least the technical improvements described above that incorporates any alleged abstract idea into a practical application thereof and/or amounts to significantly more than any alleged abstract idea. Independent claims 10 and 19 have been amended to recite similar features, and therefore the features of claims 10 and 19 provide at least the technical improvements described above that incorporate any alleged abstract idea into a practical application thereof and/or amount to significantly more than any alleged abstract idea for at least the same reasons as claim 1.
For at least these reasons, Applicant respectfully requests favorable reconsideration and withdrawal of the pending rejections under 35 U.S.C. § 101.”
In response, the examiner respectfully disagrees with Applicant’s assertions that the claimed invention has had improvements incorporated which amount to significantly more than an alleged abstract idea of a mental process. Specifically, Applicant argues two claimed improvements which are asserted to have been represented in the claims: 1. “technical solution relat[ing] to perspective-based validation of prompts to generative artificial intelligence (Al)” ([0018] of instant app, pg. 16 of remarks) and 2. “the system can execute a plurality of differently tuned or trained large language models each to generate the responses, to provide a technical improvement…” ([0022] of instant app, pg. 16 of remarks). The examiner respectfully asserts that neither of these claimed improvements have been incorporated into the claims. With regard to (1), the examiner respects Applicant’s assertions regarding a claimed improvement with respect to perspective-based validation of prompt(s) to generative artificial intelligence, but it is unclear to the examiner how this improvement can be achieved when no artificial intelligence is claimed. Further, the data being processed by the large language model is not claimed with a complexity which would prevent an interpretation of applying a mental process to a generic language model for purposes of task automation, something which does not qualify as a technical improvement (see MPEP 2106.05(a), Section I, Example iii of processes which may not be sufficient to show an improvement, “Mere automation of manual processes, such as using a generic computer to process…”). With regard to (2), the examiner respects Applicant’s assertion that the identified element of the instant app introduces a technological improvement but respectfully asserts that a plurality of differently trained or tuned models are not claimed. The claims define “a first prompt for a large language model” and “generat[ing] a plurality of second prompts for the large language model”. This is using one LLM for all prompt processing operations, not the multiple, differing models argued by Applicant. Further, there is no possible convergence or divergence calculation across a plurality of LLMs when only one is claimed.
Applicant's arguments filed 03/03/2026, see pgs. 18-21, have been fully considered but they are not persuasive.
Applicant’s representative asserts, “Zhang relates generally to ‘multi-stage processing for a large language model to answer math questions more accurately.’ (Zhang, paragraph [0017]). On pages 11-15 of the Office Action, Zhang was cited for allegedly teaching the ‘generate one or more second prompts’, ‘generate...one or more respective responses to the one or more second prompts’ and ‘cause the user interface to present...the optimized prompt or a response to the optimized prompt’ features previously recited in claim 1. Zhang discloses an ‘interface application 130 [that] may receive a natural language prompt 112 from the user 105,’ which may be ‘a natural language prompt that is related to arithmetic reasoning.’ (Zhang, paragraphs [0028]-[0029). Zhang further discloses ‘a refined prompt generator 136 configured to apply [] contextual sub-questions 142 against the original natural language prompt 112...The refined prompt generator 136 may reach a terminal state...when all of the contextual sub-questions have been applied. The refined prompt generator 136 may apply the terminal state of the refined natural language prompt to the LLM 140 and receive a final answer 148. The refined prompt generator 136 may return the final answer 148 to the user device 110.’ (Zhang, paragraph [0033]). Zhang discloses that such arrangement may allow "the interface application 130 [to] present the final answer 148 to the user 105 without presenting any intermediate prompts or answers.’ (Zhang, paragraph [0029]).
Accordingly, Zhang describes that the system refines the natural language prompt input by the user, and uses the refined prompt to generate a final answer ‘without presenting any intermediate prompts or answers.’ (See Zhang, paragraph [0029]). Therefore, Zhang does not disclose, teach, or suggest ‘generat[ing] a plurality of second prompts for the large language model based on the first prompt and the first data, each of the plurality of second prompts including one or more second data clarifying the first query’, ‘generat[ing], by the large language model receiving the plurality of second prompts, one or more respective responses to each of the plurality of second prompts,’ and ‘caus[ing] the user interface to present the one or more validated second prompts or one or more responses to the one or more validated second prompts,’ as recited in amended claim 1. (Emphasis added).
Chandrasekaran was not cited for teaching the aforementioned features of previously- recited claim 1, and as such does not cure the deficiencies of Zhang with respect to the aforementioned features of amended claim 1.
With respect to the ‘select an optimized prompt’ step previously recited in claim 1, the Examiner acknowledges, and Applicant agrees, that Zhang does not disclose ‘select an optimized prompt from among the one or more second prompts, according to a determination that a response to the at least one of the second prompt meets an accuracy threshold.’ (See Office Action, page 13). Therefore, it follows that Zhang does not disclose, teach, or suggest ‘validat[ing] one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompts meet an accuracy threshold’, and ‘select[ing] a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input,’ as recited in amended claim 1.
Chandrasekaran was instead cited to teach the ‘select an optimized prompt’ feature previously recited in claim 1. The cited portions of Chandrasekaran disclose a ‘prompt recommendation unit 110...configured to recommend enriched prompts 50 to a user based on other types of user information, such as the similarity of the stored enriched prompts 50 to prompts that were previously used or preferred by the user.’ (Chandrasekaran, paragraph [0069]). The cited portions of Chandrasekaran further disclose ‘an authorship profile matching unit 104 that can be configured to employ a text analysis technique to analyze and determine the author of [an] enriched prompt 50 and the author of [an] input prompt 18B by analyzing the language associated with each, and then match the authors of the prompts if a match exists.’ The ‘authorship profile matching unit 104 can determine and assign a matching score based on the comparison results, where the matching score is indicative of a degree of similarity between the authorship profiles of the two input and enriched prompts... [to] help quantify the degree of authorship alignment.’ (Chandrasekaran, paragraph [0074]).
Accordingly, the cited portions of Chandrasekaran appear to describe recommending an ‘enriched prompt’ based on an alignment of the ‘enriched prompt’ to an ‘input prompt.’ However, the cited portions of Chandrasekaran do not appear to disclose, teach, or suggest ‘validat[ing] one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompts meet an accuracy threshold’ and ‘select[ing] a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input,’ as recited in amended claim 1. For at least the reasons above, Applicant respectfully submits that Chandrasekaran fails to cure the deficiencies of Zhang.
For at least these reasons, amended independent claim 1 is patentable over Zhang and Chandrasekaran. Kshirsagar and Fabian are cited for allegedly teaching various additional features of claims 2-3, 5, 7, 11-12, 14, 16, and 20. Without acquiescing to these assertions, Applicant respectfully submits that Kshirsagar and Fabian fail to cure the deficiencies of Zhang and Chandrasekaran discussed above, with respect to independent claim 1.
Accordingly, and for at least the reasons discussed above, Applicant respectfully submits that independent claim 1 is patentable over each of the cited references, taken alone or in any proper combination. Independent claims 10 and 19 recite similar features to those discussed above, with respect to independent claim 1, and are thus allowable over each of the cited references, taken alone or in any proper combination, for at least similar reasons to those discussed above, with respect to independent claim 1.”
In response, with regard to Applicant’s arguments against Zhang, the examiner respectfully asserts that Applicant is arguing against elements which are not currently claimed. For example, it is unclear to the examiner how Zhang’s disclosure of generating a responses “without presenting any intermediate prompts or answers” does not teach “generat[ing] a plurality of second prompts…”, “generat[ing], by the large language model receiving the plurality of second prompts…”, and “causing the user interface to present…” when the two “generat[ing]” steps have no relation to presentation of an intermediate prompt or answer. With regard to the “caus[ing]” element, Applicant does not acknowledge the cited portion of Zhang and/or how the instant application differs from Zhang. Applicant fails to acknowledge the combination of the validated prompt of Zhang in view of the accuracy thresholding of Chandrasekaran. Further, Zhang need not discloses the “validat[ing] one or more second prompts…” element as Chandrasekaran is incorporated to resolve this deficiency of Zhang.
Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Applicant's arguments do not comply with 37 CFR 1.111(c) because they do not clearly point out the patentable novelty which he or she thinks the claims present in view of the state of the art disclosed by the references cited or the objections made. Further, they do not show how the amendments avoid such references or objections.
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
With regard to Applicant’s arguments against Chandrasekaran teaching the “validat[ing]…”, Applicant fails to disclose why the cited prompt similarity determination does not track to the prompt validation based on accuracy thresholding as currently claimed.
Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Applicant's arguments do not comply with 37 CFR 1.111(c) because they do not clearly point out the patentable novelty which he or she thinks the claims present in view of the state of the art disclosed by the references cited or the objections made. Further, they do not show how the amendments avoid such references or objections.
Applicant’s arguments, see pgs. 17-21, filed 03/03/2026, with respect to the rejection(s) of claim(s) 1, 10, and 19 under 35 U.S.C. 103 (with respect to the newly added “select[ing]” element) have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Fabian et al. (US-20240303440-A1). Fabian was used to previously reject claim 5 which contained matter similar to that which has been incorporated into the independent claims. Applicant made no arguments against the teachings of Fabian as applied to the rejection of claim 5. See updated rejections below.
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.
Claim(s) 1-6, 10-15, 20 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Independent claim(s) 1, 10, 19 recite:
receiving, via a user interface, a first prompt for a large language model including a first query that references first data;
generating a plurality of second prompts for the large language model based on the first prompt and the first data, each of the plurality of second prompts including one or more second data clarifying the first query;
generating, by the large language model receiving the plurality of second prompts, one or more respective responses to each of the plurality of second prompts;
validating one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompts meets an accuracy threshold;
causing the user interface to present the one or more validated second prompts or one or more responses to the one or more validated second prompts; and
selecting a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input.
These limitations, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. For example, the claim(s) read(s) on receiving a text for prompting a response to a query based on associated data of a query, i.e. all written information, generating additional written prompts to be asked for clarification purposes, wherein that text contains second data, i.e. written entities, generating a response to the written prompt, selecting a prompt which best answers the required query based on an accuracy measure, and presenting the optimized prompt and/or presenting a generated response from the optimized prompt. That is, other than reciting “large language model” (claims 1, 10, 19) and/or processor(s) and memory (claims 1, 19), nothing in the claim element precludes the step from practically being performed in the mind.
For example, the step of “receiving, via a user interface, a first prompt…including a first query that references first data”, all of these elements can be written and or mentally monitored with the aid of pen and paper. Generating second prompts based on the first prompt and first data is equivalent to writing down additional questions/information related to the first data as would be mentally determined, i.e. if the prompt is hotel reservation in Chicago, an additional prompt would be “What area of the city are you looking to stay in?”, which is based on the original first data of “location = Chicago” with clarifying information, i.e. neighborhood determination step in the prompt. Generating responses to prompts containing textual queries is a mental process aided with pen and paper, i.e. the reader of the prompt can mentally determine responses. Validating a prompt based on an accuracy threshold is equivalent to comparing the ordering and/or numbers of words between two texts, something which can be performed mentally with the aid of pen and paper. Further, as previously disclosed, generating responses to queries/prompts is a mental process aided with pen and paper. Further, a written document can be provided to a user, wherein their brain is effectively a user interface for processing/perceiving information, i.e. the written prompts/responses. Lastly, selecting a prompt via a user interface for response generation is an operation equivalent to presenting the user all prompts written on paper to be selected by the user based on a reading of the prompts. All of these steps can be performed in the mind and/or using pen and paper.
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea (Step 2A, Prong one, Yes).
This judicial exception is not integrated into a practical application because the addition of generically recited computer elements does not add a meaningful limitation to the abstract idea because they amount to simply implementing the abstract idea on a computer. The claims are directed to an abstract idea. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception ( Step 2A, Prong two, No). As discussed above, with respect to integration of an abstract idea into a practical application, the additional element of “receiving”, “generating”, “storing”, “presenting” are merely for the purpose of data gathering, storing, processing, and/or insignificant extra-solution activity that amount to no more than mere instructions to apply the exception using a generic computer component. Paragraph(s) [0025] (with respect to the computing system), [0037] (with respect to the client device) of the instant application disclose(s) applying the method to a generic computing device such as a PC. Mere instructions to apply the exception using a generic computer component cannot provide an inventive concept. Therefore, the claims are not patent eligible (Step 2B, No).
Similarly, dependent claim(s) 2-6, 11-15, 20 include additional steps that are considered “insignificant extra-solution activity to judicial exception” because they fail to provide meaningful significance that goes beyond generally linking the use of an abstract idea to a particular technological environment.
For example, claim 2 reads on filtering restricted information from a first prompt to result in a second prompt also containing second data. Removing restricted, i.e. personal/private, information can be performed mentally through reading the prompt, aided by pen and paper, with a written document. Determining to add second data to clarify a first query given the first query is also a mental process associated with writing when more information is needed to answer the question, as would be determined mentally.
Claim 3 reads on generating a third prompt including a second query to clarify the first query and first data. As previously disclosed, prompt generation is a mental process assisted with pen and paper. Generation of queries is similarly a mental process assisted with pen and paper. Extending this operation to a third prompt/response does not prevent the operation from being performed mentally.
Claim 4 reads on each of the second prompts having different second data. Determining the type of data to add to a written prompt, i.e. also writing, is a mental process assisted with pen and paper.
Claim 5 reads on presenting the second prompts, obtaining a selected second prompt, and selecting that prompt to be the optimized prompt, wherein the selected prompt is used to generate a response. As previously disclosed, prompts can be written; therefore, presentation of a written document for selection is a mental process. Selecting one block of text, i.e. prompt, given a plurality is a mental process. Further, one could write “optimized” next to the selected prompt as a method of “selecting”. Generating a response is a mental process associated with reading/thinking based on a received prompt.
Claim 6 reads on the accuracy threshold determination being based upon differences in words in responses. Comparing words between responses is a mental process associated with reading the words.
Claim 11 reads on filtering restricted information from a first prompt to result in a second prompt also containing second data. Removing restricted, i.e. personal/private, information can be performed mentally through reading the prompt, aided by pen and paper, with a written document. Determining to add second data to clarify a first query given the first query is also a mental process associated with writing when more information is needed to answer the question, as would be determined mentally.
Claim 12 reads on generating a third prompt including a second query to clarify the first query and first data. As previously disclosed, prompt generation is a mental process assisted with pen and paper. Generation of queries is similarly a mental process assisted with pen and paper. Extending this operation to a third prompt/response does not prevent the operation from being performed mentally.
Claim 13 reads on each of the second prompts having different second data. Determining the type of data to add to a written prompt, i.e. also writing, is a mental process assisted with pen and paper.
Claim 14 reads on presenting the second prompts, obtaining a selected second prompt, and selecting that prompt to be the optimized prompt, wherein the selected prompt is used to generate a response. As previously disclosed, prompts can be written; therefore, presentation of a written document for selection is a mental process. Selecting one block of text, i.e. prompt, given a plurality is a mental process. Further, one could write “optimized” next to the selected prompt as a method of “selecting”. Generating a response is a mental process associated with reading/thinking based on a received prompt.
Claim 15 reads on the accuracy threshold determination being based upon differences in words in responses. Comparing words between responses is a mental process associated with reading the words.
Claim 20 reads on generating a third prompt including a second query to clarify the first query and first data. As previously disclosed, prompt generation is a mental process assisted with pen and paper. Generation of queries is similarly a mental process assisted with pen and paper. Extending this operation to a third prompt/response does not prevent the operation from being performed mentally.
Therefore, these claims are also not patent eligible.
Claims 7 and 16 have been deemed to be containing eligible subject matter as the disclosure that the large language model is being performed on a computing device in a system with two distinct computing devices incorporates an amount of structure into the claimed invention which precludes an interpretation of the claimed operations from being performed mentally with the aid of pen and paper.
Claims 8 and 17 have been deemed to be containing eligible subject matter as the disclosure that the prompts are transmitted to a search engine for response generation indicates that this operations cannot be reasonably performed in the mind as a search engine is a specific software querying structure.
Claims 9 and 18 have been deemed to be containing eligible subject matter as the disclosure that the prompt is used to generate a non-fungible token (NFT) is something which cannot be reasonably performed in the mind as NFTs are digital in nature.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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.
Claim(s) 1, 4-6, 8-10, 13-15, 17-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Zhang (US-20240370658-A1) in view of Chandrasekaran (US-20240320476-A1), further in view of Fabian et al. (US-20240303440-A1), hereinafter Fabian.
Regarding claim 1, Zhang discloses: a system ([Fig. 1, system 120]), comprising:
one or more processing circuits comprising memory storing instructions therein that are executable by one or more processors ([0040] In an example, the apparatus 300 includes at least one processor 302 and a memory 304 configured to execute or store instructions) to cause the one or more processors to:
receive, via a user interface ([0028] interface application 130 may provide a graphical user interface on the user device 110 for the user 105), a first prompt for a large language model including a first query that references first data ([Fig. 1, Large Language Model 140 receiving natural language prompt 112 as part of datacenter 122], [0028] The interface application 130 may receive a natural language prompt 112 from the user 105, [0029] the interface application 130 is configured to refine a natural language prompt that is related to arithmetic reasoning, [0017] multi-stage processing for a large language model to answer math questions more accurately, [0030] a question related to arithmetic reasoning may include at least numerical values and an operation or comparison, [Answering math questions indicates a received questions, i.e. query, with first data, i.e. numerical values and/or operations related to the arithmetic]);
generate a plurality of second prompts for the large language model based on the first prompt and the first data ([0033] The interface application 130 includes a refined prompt generator 136 configured to apply the contextual sub-questions 142 against the original natural language prompt 112 with the contextual answers 144 as a refined natural language prompt 146 to the LLM 140 in a reverse order of the series of sub-questions 142. The refined prompt generator 136 may provide the refined prompts 146 to the LLM 140, [A refined prompt resulting through the application of contextual sub-questions against an original prompt (containing the first query and associated first data) indicates the refined prompt to be based on the first prompt and first data]), each of the plurality of second prompts including one or more second data clarifying the first query ([0033] For instance, the sub-question generator 134 may call the sub-question generator 134 for a further sub-question and/or the refined prompt generator 134 when a contextual answer is received. In some implementations, the sub-question generator 134 may add sub-questions to a stack, and the refined prompt generator 134 may process answers to generate a refined prompt, [Contextual sub-questions with associated contextual answers being processed to refine the prompt indicates the contextual answers to be second data clarifying the first query resulting in the refined, i.e. second, prompt]); and,
generate, by the large language model receiving the plurality of second prompts ([Fig. 1, Large Language Model 140 receiving Refined Prompts 146]), one or more respective responses to each of the plurality of second prompts ([Fig. 1, Final Answer 148 output from LLM 140], [0033] The refined prompt generator 136 may apply the terminal state of the refined natural language prompt to the LLM 140 and receive a final answer 148).
Zhang does not disclose:
validate one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompts meets an accuracy threshold.
Chandrasekaran discloses:
validate one or more second prompts from among the plurality of second prompts ([0033] As used herein, the term “enrich,” “enriched” or “enriching” is intended to include the ability to ingest, integrate, augment, improve and/or enhance data by supplementing missing or incomplete data, correcting inaccurate data, [0073] The trending prompt unit 102 can optionally rank a plurality of the enriched prompts based on the popularity attributes associated therewith, and the higher ranked prompts can be selected as trending prompts, [An enriched prompt tracks to a second, i.e. refined, prompt as previously disclosed by Zhang. Further, generating a plurality of enriched prompts, wherein the highest ranked prompts are selected, indicates the selected top-ranked prompts to be validated as compared to those which aren’t selected]), according to a determination that a response to each of the one or more validated second prompts meets an accuracy threshold ([0061] The enriched prompt 50 generated by the prompt enrichment unit 20 can also be conveyed to a prompt filtering unit 60 for performing one or more filtering operations or techniques on the language in the enriched prompt 50. The prompt filtering unit 60 can filter and detect for the presence of certain language within the enriched prompt 50 and determine if the language in the enriched prompt 50 previously existed, [0069] The similarities between the prompts can be determined based on the writing style within the prompts, the presence of certain keywords or similar words in prompts, [0074] Further, depending on the specific application, a threshold score can be applied to determine whether the prompts are deemed to have a sufficiently similar authorship profile…If the similarity score exceeds the threshold score or value, the prompts may be considered to have matching authorship profiles, [Filtering prompts based on the language within a prompt as compared to previous language indicates the filtering to be based on an accuracy of language between the enriched prompt and previous data. Further, determining common authorship between two prompts based on similarity of prompts, wherein prompts are comprised of words, indicates the similarity comparison to be accuracy based as compared to previously entered data. Further still, if exceeding a similarity threshold indicates a prompt match, this indicates the similarity to be word accuracy-based]).
Zhang and Chandrasekaran are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang to incorporate the teachings of Chandrasekaran, because of the novel way to enrich a prompt to increase the ease in which a subsequent user will be able to determine whether or not that prompt is relevant and/or effective for their intended purposes (Chandrasekaran, [0008]).
Zhang further discloses:
cause the user interface to present the one or more validated second prompts ([The examiner would like to note that due to the disjunctive nature of this element, presentation of the optimized prompt is not a required element; therefore, no mapping has been provided]) or one or more responses to the one or more validated second prompts ([0029] the interface application 130 may present the final answer 148 to the user 105).
Zhang in view of Chandrasekaran does not disclose:
select a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input.
Fabian discloses:
select a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input ([Fig. 5F, User Interface 307], [Fig. 5F, LLM Generated Suggestions, Adding a “Last name” column on the user interface], [Wherein the suggestion to add a name column is part of at least a second prompt in view of the preceding conversation steps in Figs. 5B-5F, wherein the option to add the column is necessarily performed by clicking the “Add column” button via user interface, the updated display represents the generated response], [Fig. 5F, Receiving Instructions from the Prompt Engine 305 resulting in an updated display], [Instructions to be implemented on an application component indicates the instructions to be a prompt for the application component(s) to update a visual layout. Further, the examiner would like to note that if only one second prompt is selected, as indicated to be possibly occurring in the previous claim element, then the selected second prompt is the optimized prompt automatically]).
Zhang, Chandrasekaran, and Fabian are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran to incorporate the teachings of Fabian, because of the novel way to interpret received user inquiries in multiple ways and to generate suggestions based on each interpretation, wherein follow-up prompts are generated based on the user suggestion, improving the focus of replies or suggestions generated by an LLM (Fabian, [0027]).
Regarding claim 4, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Zhang further discloses:
the one or more processors further configured to:
generate each of the plurality of second prompts having a different one of the one or more second data ([Fig. 2, Sub question 212 -Sub question n 218], [0021] refining the original natural language prompt based on the contextual sub-questions and contextual answers, [Generating a plurality of sub-questions, all used to generate the refined, i.e. second, prompt, indicates the second prompts to be all having different second data based on the sub-question(s) asked (context dependency indicates different contexts to have different associated data)]).
Regarding claim 5, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Chandrasekaran further discloses:
the one or more processors further configured to:
present, via the user interface ([0037] a user interface), the one or more validated second prompts ([0070] configured to recommend enriched prompts 50 to a user based on other selected types of user information, [Enriched prompts track to second, i.e. refined, prompts as previously disclosed by Zhang]).
Fabian further discloses:
the one or more processors further configured to:
obtain, via the user interface ([Fig. 5F, User Interface 307]), a selection of one or more of the validated second prompts ([Fig. 5F, LLM Generated Suggestions, Adding a “Last name” column on the user interface], [Wherein the suggestion to add a name column is part of at least a second prompt in view of the preceding conversation steps in Figs. 5B-5F]);
select an optimized prompt from among a subset of the one or more of the validated second prompts selected via the user interface ([Fig. 5F, Receiving Instructions from the Prompt Engine 305 resulting in an updated display], [Instructions to be implemented on an application component indicates the instructions to be a prompt for the application component(s) to update a visual layout. Further, the examiner would like to note that if only one second prompt is selected, as indicated to be possibly occurring in the previous claim element, then the selected second prompt is the optimized prompt automatically]), the optimized prompt corresponding to the selected validated second prompt used to generate the response ([A user selected update to a display indicates that the optimization (selected by the user) corresponds to a validated second prompt for generating a response, otherwise the optimization would not be chosen, i.e. validated]).
Regarding claim 6, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Chandrasekaran further discloses:
the one or more processors further configured to:
determine that the response meets the accuracy threshold based on a difference between presence of one or more words in the response and one or more words in each of the responses ([0051] An accurate output is one that is correct in all included details and that is presented using appropriate syntax, [0069] configured to recommend enriched prompts 50 to a user based on other types of user information, such as the similarity of the stored enriched prompts 50 to prompts that were previously used or preferred by the user. The similarities between the prompts can be determined based on the writing style within the prompts, the presence of certain keywords or similar words in prompts, similar contexts, similar author or provider attributes, similar classifications, and the like, [Syntax represents word ordering indicating that an accurate output will inherently have a dependence on word ordering indicating that the syntax of a response would be affected by different words between responses. Further, recommending prompts based on a condition of similar words indicates a threshold similarity/accuracy between words for presentation]);
wherein the accuracy threshold is a maximum difference between words among the responses ([As previously disclosed, determining a similarity between prompts (prompts containing responses as would be required for the step of seeing if a prompt meets the accuracy threshold of claim 1, i.e. if the same accuracy measure can be applied to prompts (claim 1) and responses (claim 6), this indicates the prompts to be containing responses and/or consisting of responses) based on a word similarity indicates the word similarity to be indicative of an accuracy, wherein using this metric as a means for presentation determination further indicates a maximum threshold difference between words for determining which enriched prompts are presented]).
Regarding claim 8, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Zhang further discloses:
the one or more processors further configured to:
transmit, to a search engine, one or more of the second prompts ([0019] In an aspect, the present disclosure provides techniques for an interface between a user and a large language model to use multiple queries to refine a natural language prompt from the user before returning a final answer to the user. For example, the interface may be a search engine, [Any additional refinement after a first refinement, see Fig. 7, 710 “second level refined natural language prompt”, tracks to performing the cited operation on a second prompt, necessarily requiring a transmission of the prompt to the search engine to receive answers]); and,
obtain, from the search engine ([In view of the previously disclosed search engine]), one or more responses to the second prompts ([Fig. 2, Answer from Updated Question 224], [An updated question indicates the prompt containing that question to be refined, i.e. second, as compared to the original question 212. Further, as Zhang discloses using a search engine for queries, it would not be extending beyond the disclosure of Zhang to suggest that a search engine could be used to generate the answers of Fig. 2]).
Chandrasekaran further discloses:
determine, based on one or more responses to the second prompts, that the response to the at least one of the second prompts meets the accuracy threshold ([0061] The enriched prompt 50 generated by the prompt enrichment unit 20 can also be conveyed to a prompt filtering unit 60 for performing one or more filtering operations or techniques on the language in the enriched prompt 50. The prompt filtering unit 60 can filter and detect for the presence of certain language within the enriched prompt 50 and determine if the language in the enriched prompt 50 previously existed, [0074] Further, depending on the specific application, a threshold score can be applied to determine whether the prompts are deemed to have a sufficiently similar authorship profile…If the similarity score exceeds the threshold score or value, the prompts may be considered to have matching authorship profiles, [Filtering prompts based on the language within a prompt as compared to previous language indicates the filtering to be based on an accuracy of language between the enriched prompt and previous data. Further, determining common authorship between two prompts based on similarity of prompts, wherein prompts are comprised of words, indicates the similarity comparison to be word-accuracy based as compared to previously entered data]).
Regarding claim 9, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Chandrasekaran further discloses:
the one or more processors further configured to:
generate a non-fungible token (NFT) based on an optimized prompt ([0056] convert the ontology prompt 28 or an enriched prompt 50 into a non-fungible token (NFT)), the optimized prompt corresponding to the selected validated second prompt ([An enriched prompt as previously cited tracks to an optimized prompt]); and,
cause a blockchain provider system to register the NFT for the optimized prompt to a blockchain for one or more optimized prompts ([0076] employ blockchain technology to ensure tamper-proof lineage and veracity of stored prompts, establishing a secure and immutable record of prompt creation, enrichment, and modification history. Each prompt stored in the blockchain ledger can be associated with one or more Non-Fungible Tokens (NFTs), endowing individual prompts with unique identifiers that certify their authenticity, ownership, and originality).
Regarding claim 10, Zhang discloses: a method ([0044] a method 400), comprising:
receiving, via a user interface ([0028] interface application 130 may provide a graphical user interface on the user device 110 for the user 105), a first prompt for a large language model including a first query that references first data ([Fig. 1, Large Language Model 140 receiving natural language prompt 112 as part of datacenter 122], [0028] The interface application 130 may receive a natural language prompt 112 from the user 105, [0029] the interface application 130 is configured to refine a natural language prompt that is related to arithmetic reasoning, [0017] multi-stage processing for a large language model to answer math questions more accurately, [0030] a question related to arithmetic reasoning may include at least numerical values and an operation or comparison, [Answering math questions indicates a received questions, i.e. query, with first data, i.e. numerical values and/or operations related to the arithmetic]);
generating a plurality of second prompts for the large language model based on the first prompt and the first data ([0033] The interface application 130 includes a refined prompt generator 136 configured to apply the contextual sub-questions 142 against the original natural language prompt 112 with the contextual answers 144 as a refined natural language prompt 146 to the LLM 140 in a reverse order of the series of sub-questions 142. The refined prompt generator 136 may provide the refined prompts 146 to the LLM 140, [A refined prompt resulting through the application of contextual sub-questions against an original prompt (containing the first query and associated first data) indicates the refined prompt to be based on the first prompt and first data]), each of the plurality of second prompts including one or more second data clarifying the first query ([0033] For instance, the sub-question generator 134 may call the sub-question generator 134 for a further sub-question and/or the refined prompt generator 134 when a contextual answer is received. In some implementations, the sub-question generator 134 may add sub-questions to a stack, and the refined prompt generator 134 may process answers to generate a refined prompt, [Contextual sub-questions with associated contextual answers being processed to refine the prompt indicates the contextual answers to be second data clarifying the first query resulting in the refined, i.e. second, prompt]); and,
generating, by the large language model receiving the plurality of second prompts ([Fig. 1, Large Language Model 140 receiving Refined Prompts 146]), one or more respective responses to each of the plurality of second prompts ([Fig. 1, Final Answer 148 output from LLM 140], [0033] The refined prompt generator 136 may apply the terminal state of the refined natural language prompt to the LLM 140 and receive a final answer 148).
Zhang does not disclose:
validating one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompt meets an accuracy threshold.
Chandrasekaran discloses:
validating one or more second prompts from among the plurality of second prompts ([0033] As used herein, the term “enrich,” “enriched” or “enriching” is intended to include the ability to ingest, integrate, augment, improve and/or enhance data by supplementing missing or incomplete data, correcting inaccurate data, [0073] The trending prompt unit 102 can optionally rank a plurality of the enriched prompts based on the popularity attributes associated therewith, and the higher ranked prompts can be selected as trending prompts, [An enriched prompt tracks to a second, i.e. refined, prompt as previously disclosed by Zhang. Further, generating a plurality of enriched prompts, wherein the highest ranked prompts are selected, indicates the selected top-ranked prompts to be validated as compared to those which aren’t selected]), according to a determination that a response to each of the one or more validated second prompt meets an accuracy threshold ([0061] The enriched prompt 50 generated by the prompt enrichment unit 20 can also be conveyed to a prompt filtering unit 60 for performing one or more filtering operations or techniques on the language in the enriched prompt 50. The prompt filtering unit 60 can filter and detect for the presence of certain language within the enriched prompt 50 and determine if the language in the enriched prompt 50 previously existed, [0069] The similarities between the prompts can be determined based on the writing style within the prompts, the presence of certain keywords or similar words in prompts, [0074] Further, depending on the specific application, a threshold score can be applied to determine whether the prompts are deemed to have a sufficiently similar authorship profile…If the similarity score exceeds the threshold score or value, the prompts may be considered to have matching authorship profiles, [Filtering prompts based on the language within a prompt as compared to previous language indicates the filtering to be based on an accuracy of language between the enriched prompt and previous data. Further, determining common authorship between two prompts based on similarity of prompts, wherein prompts are comprised of words, indicates the similarity comparison to be accuracy based as compared to previously entered data. Further still, if exceeding a similarity threshold indicates a prompt match, this indicates the similarity to be word accuracy-based]).
Zhang and Chandrasekaran are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang to incorporate the teachings of Chandrasekaran, because of the novel way to enrich a prompt to increase the ease in which a subsequent user will be able to determine whether or not that prompt is relevant and/or effective for their intended purposes (Chandrasekaran, [0008]).
Zhang further discloses:
causing the user interface to present the one or more validated second prompts ([The examiner would like to note that due to the disjunctive nature of this element, presentation of the optimized prompt is not a required element; therefore, no mapping has been provided]) or one or more responses to the one or more validated second prompts ([0029] the interface application 130 may present the final answer 148 to the user 105).
Zhang in view of Chandrasekaran does not disclose:
selecting a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input.
Fabian discloses:
selecting a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input ([Fig. 5F, User Interface 307], [Fig. 5F, LLM Generated Suggestions, Adding a “Last name” column on the user interface], [Wherein the suggestion to add a name column is part of at least a second prompt in view of the preceding conversation steps in Figs. 5B-5F, wherein the option to add the column is necessarily performed by clicking the “Add column” button via user interface, the updated display represents the generated response], [Fig. 5F, Receiving Instructions from the Prompt Engine 305 resulting in an updated display], [Instructions to be implemented on an application component indicates the instructions to be a prompt for the application component(s) to update a visual layout. Further, the examiner would like to note that if only one second prompt is selected, as indicated to be possibly occurring in the previous claim element, then the selected second prompt is the optimized prompt automatically]).
Zhang, Chandrasekaran, and Fabian are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran to incorporate the teachings of Fabian, because of the novel way to interpret received user inquiries in multiple ways and to generate suggestions based on each interpretation, wherein follow-up prompts are generated based on the user suggestion, improving the focus of replies or suggestions generated by an LLM (Fabian, [0027]).
Regarding claim 13, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Zhang further discloses:
generating each of the plurality of second prompts having a different one of the one or more second data ([Fig. 2, Sub question 212 -Sub question n 218], [0021] refining the original natural language prompt based on the contextual sub-questions and contextual answers, [Generating a plurality of sub-questions, all used to generate the refined, i.e. second, prompt, indicates the second prompts to be all having different second data based on the sub-question(s) asked (context dependency indicates different contexts to have different associated data)]).
Regarding claim 14, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Chandrasekaran further discloses:
presenting, via the user interface ([0037] a user interface), the one or more validated second prompts ([0070] configured to recommend enriched prompts 50 to a user based on other selected types of user information, [Enriched prompts track to second, i.e. refined, prompts as previously disclosed by Zhang]).
Fabian further discloses:
obtaining, via the user interface ([Fig. 5F, User Interface 307]), a selection of one or more of the validated second prompts ([Fig. 5F, LLM Generated Suggestions, Adding a “Last name” column on the user interface], [Wherein the suggestion to add a name column is part of at least a second prompt in view of the preceding conversation steps in Figs. 5B-5F]);
selecting an optimized prompt from among a subset of the one or more of the validated second prompts selected via the user interface ([Fig. 5F, Receiving Instructions from the Prompt Engine 305 resulting in an updated display], [Instructions to be implemented on an application component indicates the instructions to be a prompt for the application component(s) to update a visual layout. Further, the examiner would like to note that if only one second prompt is selected, as indicated to be possibly occurring in the previous claim element, then the selected second prompt is the optimized prompt automatically])., the optimized prompt corresponding to the selected validated second prompt used to generate the response ([A user selected update to a display indicates that the optimization (selected by the user) corresponds to a validated second prompt for generating a response, otherwise the optimization would not be chosen, i.e. validated]).
Regarding claim 15, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Chandrasekaran further discloses:
determining that the response meets the accuracy threshold based on a difference between presence of one or more words in the response and one or more words in each of the responses ([0051] An accurate output is one that is correct in all included details and that is presented using appropriate syntax, [0069] configured to recommend enriched prompts 50 to a user based on other types of user information, such as the similarity of the stored enriched prompts 50 to prompts that were previously used or preferred by the user. The similarities between the prompts can be determined based on the writing style within the prompts, the presence of certain keywords or similar words in prompts, similar contexts, similar author or provider attributes, similar classifications, and the like, [Syntax represents word ordering indicating that an accurate output will inherently have a dependence on word ordering indicating that the syntax of a response would be affected by different words between responses. Further, recommending prompts based on a condition of similar words indicates a threshold similarity/accuracy between words for presentation]);
wherein the accuracy threshold is a maximum difference between words among the responses ([As previously disclosed, determining a similarity between prompts (prompts containing responses as would be required for the step of seeing if a prompt meets the accuracy threshold of claim 1, i.e. if the same accuracy measure can be applied to prompts (claim 1) and responses (claim 6), this indicates the prompts to be containing responses and/or consisting of responses) based on a word similarity indicates the word similarity to be indicative of an accuracy, wherein using this metric as a means for presentation determination further indicates a maximum threshold difference between words for determining which enriched prompts are presented]).
Regarding claim 17, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Zhang further discloses:
transmitting, to a search engine, one or more of the second prompts ([0019] In an aspect, the present disclosure provides techniques for an interface between a user and a large language model to use multiple queries to refine a natural language prompt from the user before returning a final answer to the user. For example, the interface may be a search engine, [Any additional refinement after a first refinement, see Fig. 7, 710 “second level refined natural language prompt”, tracks to performing the cited operation on a second prompt, necessarily requiring a transmission of the prompt to the search engine to receive answers]); and,
obtaining, from the search engine ([In view of the previously disclosed search engine]), one or more responses to the second prompts ([Fig. 2, Answer from Updated Question 224], [An updated question indicates the prompt containing that question to be refined, i.e. second, as compared to the original question 212. Further, as Zhang discloses using a search engine for queries, it would not be extending beyond the disclosure of Zhang to suggest that a search engine could be used to generate the answers of Fig. 2]).
Chandrasekaran further discloses:
determining, based on one or more responses to the second prompts, that the response to the at least one of the second prompts meets the accuracy threshold ([0061] The enriched prompt 50 generated by the prompt enrichment unit 20 can also be conveyed to a prompt filtering unit 60 for performing one or more filtering operations or techniques on the language in the enriched prompt 50. The prompt filtering unit 60 can filter and detect for the presence of certain language within the enriched prompt 50 and determine if the language in the enriched prompt 50 previously existed, [0074] Further, depending on the specific application, a threshold score can be applied to determine whether the prompts are deemed to have a sufficiently similar authorship profile…If the similarity score exceeds the threshold score or value, the prompts may be considered to have matching authorship profiles, [Filtering prompts based on the language within a prompt as compared to previous language indicates the filtering to be based on an accuracy of language between the enriched prompt and previous data. Further, determining common authorship between two prompts based on similarity of prompts, wherein prompts are comprised of words, indicates the similarity comparison to be word-accuracy based as compared to previously entered data]).
Regarding claim 18, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Chandrasekaran further discloses:
generate a non-fungible token (NFT) based on the optimized prompt ([0056] convert the ontology prompt 28 or an enriched prompt 50 into a non-fungible token (NFT)), the optimized prompt corresponding to the selected validated second prompt ([An enriched prompt as previously cited tracks to an optimized prompt]); and,
cause a blockchain provider system to register the NFT for the optimized prompt to a blockchain for one or more optimized prompts ([0076] employ blockchain technology to ensure tamper-proof lineage and veracity of stored prompts, establishing a secure and immutable record of prompt creation, enrichment, and modification history. Each prompt stored in the blockchain ledger can be associated with one or more Non-Fungible Tokens (NFTs), endowing individual prompts with unique identifiers that certify their authenticity, ownership, and originality).
Regarding claim 19, Zhang discloses: a non-transitory computer readable medium including one or more instructions stored thereon an executable by a processor to ([0040] In an example, the apparatus 300 includes at least one processor 302 and a memory 304 configured to execute or store instructions, [0071] Non-transitory computer-readable media):
receive, via a user interface ([0028] interface application 130 may provide a graphical user interface on the user device 110 for the user 105), a first prompt for a large language model including a first query that references first data ([Fig. 1, Large Language Model 140 receiving natural language prompt 112 as part of datacenter 122], [0028] The interface application 130 may receive a natural language prompt 112 from the user 105, [0029] the interface application 130 is configured to refine a natural language prompt that is related to arithmetic reasoning, [0017] multi-stage processing for a large language model to answer math questions more accurately, [0030] a question related to arithmetic reasoning may include at least numerical values and an operation or comparison, [Answering math questions indicates a received questions, i.e. query, with first data, i.e. numerical values and/or operations related to the arithmetic]);
generate a plurality of second prompts for the large language model based on the first prompt and the first data ([0033] The interface application 130 includes a refined prompt generator 136 configured to apply the contextual sub-questions 142 against the original natural language prompt 112 with the contextual answers 144 as a refined natural language prompt 146 to the LLM 140 in a reverse order of the series of sub-questions 142. The refined prompt generator 136 may provide the refined prompts 146 to the LLM 140, [A refined prompt resulting through the application of contextual sub-questions against an original prompt (containing the first query and associated first data) indicates the refined prompt to be based on the first prompt and first data]), each of the plurality of second prompts including one or more second data clarifying the first query ([0033] For instance, the sub-question generator 134 may call the sub-question generator 134 for a further sub-question and/or the refined prompt generator 134 when a contextual answer is received. In some implementations, the sub-question generator 134 may add sub-questions to a stack, and the refined prompt generator 134 may process answers to generate a refined prompt, [Contextual sub-questions with associated contextual answers being processed to refine the prompt indicates the contextual answers to be second data clarifying the first query resulting in the refined, i.e. second, prompt]); and,
generate, by the large language model receiving the plurality of second prompts ([Fig. 1, Large Language Model 140 receiving Refined Prompts 146]), one or more respective responses to each of the plurality of second prompts ([Fig. 1, Final Answer 148 output from LLM 140], [0033] The refined prompt generator 136 may apply the terminal state of the refined natural language prompt to the LLM 140 and receive a final answer 148).
Zhang does not disclose:
validate one or more second prompts from among the plurality of second prompts, according to a determination that a response to each of the one or more validated second prompt meets an accuracy threshold.
Chandrasekaran discloses:
validate one or more second prompts from among the plurality of second prompts ([0033] As used herein, the term “enrich,” “enriched” or “enriching” is intended to include the ability to ingest, integrate, augment, improve and/or enhance data by supplementing missing or incomplete data, correcting inaccurate data, [0073] The trending prompt unit 102 can optionally rank a plurality of the enriched prompts based on the popularity attributes associated therewith, and the higher ranked prompts can be selected as trending prompts, [An enriched prompt tracks to a second, i.e. refined, prompt as previously disclosed by Zhang. Further, generating a plurality of enriched prompts, wherein the highest ranked prompts are selected, indicates the selected top-ranked prompts to be validated as compared to those which aren’t selected]), according to a determination that a response to each of the one or more validated second prompt meets an accuracy threshold ([0061] The enriched prompt 50 generated by the prompt enrichment unit 20 can also be conveyed to a prompt filtering unit 60 for performing one or more filtering operations or techniques on the language in the enriched prompt 50. The prompt filtering unit 60 can filter and detect for the presence of certain language within the enriched prompt 50 and determine if the language in the enriched prompt 50 previously existed, [0069] The similarities between the prompts can be determined based on the writing style within the prompts, the presence of certain keywords or similar words in prompts, [0074] Further, depending on the specific application, a threshold score can be applied to determine whether the prompts are deemed to have a sufficiently similar authorship profile…If the similarity score exceeds the threshold score or value, the prompts may be considered to have matching authorship profiles, [Filtering prompts based on the language within a prompt as compared to previous language indicates the filtering to be based on an accuracy of language between the enriched prompt and previous data. Further, determining common authorship between two prompts based on similarity of prompts, wherein prompts are comprised of words, indicates the similarity comparison to be accuracy based as compared to previously entered data. Further still, if exceeding a similarity threshold indicates a prompt match, this indicates the similarity to be word accuracy-based]).
Zhang and Chandrasekaran are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang to incorporate the teachings of Chandrasekaran, because of the novel way to enrich a prompt to increase the ease in which a subsequent user will be able to determine whether or not that prompt is relevant and/or effective for their intended purposes (Chandrasekaran, [0008]).
Zhang further discloses:
cause the user interface to present the one or more validated second prompts ([The examiner would like to note that due to the disjunctive nature of this element, presentation of the optimized prompt is not a required element; therefore, no mapping has been provided]) or one or more responses to the one or more validated second prompts ([0029] the interface application 130 may present the final answer 148 to the user 105).
Zhang in view of Chandrasekaran does not disclose:
select a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input.
Fabian discloses:
select a validated second prompt of the one or more validated second prompts based on an input received via the user interface, the large language model configured to generate a response using the selected validated second prompt as input ([Fig. 5F, User Interface 307], [Fig. 5F, LLM Generated Suggestions, Adding a “Last name” column on the user interface], [Wherein the suggestion to add a name column is part of at least a second prompt in view of the preceding conversation steps in Figs. 5B-5F, wherein the option to add the column is necessarily performed by clicking the “Add column” button via user interface, the updated display represents the generated response], [Fig. 5F, Receiving Instructions from the Prompt Engine 305 resulting in an updated display], [Instructions to be implemented on an application component indicates the instructions to be a prompt for the application component(s) to update a visual layout. Further, the examiner would like to note that if only one second prompt is selected, as indicated to be possibly occurring in the previous claim element, then the selected second prompt is the optimized prompt automatically]).
Zhang, Chandrasekaran, and Fabian are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran to incorporate the teachings of Fabian, because of the novel way to interpret received user inquiries in multiple ways and to generate suggestions based on each interpretation, wherein follow-up prompts are generated based on the user suggestion, improving the focus of replies or suggestions generated by an LLM (Fabian, [0027]).
Claim(s) 2-3, 7, 11-12, 16, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Zhang in view of Chandrasekaran, further in view of Fabian, further in view of Kshirsagar et al. (US-20250272510-A1), hereinafter Kshirsagar.
Regarding claim 2, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
the one or more processors further configured to:
filter the first data from the first prompt into the plurality of second prompts, each of the plurality of second prompts excluding the first data and including one or more second data clarifying the first query,
wherein the first data is restricted from transmission.
Kshirsagar discloses:
the one or more processors further configured to:
filter the first data from the first prompt into the plurality of second prompts ([0141] Upon determining to mask sensitive information, sensitive information in the prompt is identified and replaced with unique identifiers at 812, [0148] Upon determining to replace sensitive information, the unique identifiers added to the prompt at 812 may be replaced with the corresponding sensitive information at 824, [Adding sensitive information back into a prompt after prompt completion, see Fig. 8, 816, indicates a filtering of the sensitive data from the first prompt, i.e. originally masked/incomplete, back into the second, i.e. complete, prompt]), each of the plurality of second prompts excluding the first data and including one or more second data clarifying the first query ([It is unclear to the examiner how the second prompts can possibly be excluding the first data when the first data has been previously filtered into the second prompts. Further, consider the unique identifiers associated with sensitive information to be first data associated with a first prompt/query after filtering and the replacement of these identifiers with the sensitive information to be second data clarifying the first query]),
wherein the first data is restricted from transmission ([Fig. 8, 822 Replace sensitive information?, No], [Opting to not replace identifiers with sensitive information indicates the sensitive information to be restricted from transmission]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 3, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
the one or more processors further configured to:
generate, by the large language model receiving the first prompt, a third prompt for a user including a second query to clarify at least one of the first query and first data;
transmit, to the user interface, the third prompt;
obtain, via the user interface, a response to the third prompt, the response to the third prompt including the second data clarifying at least one of the first query and the first data.
Kshirsagar discloses:
the one or more processors further configured to:
generate, by the large language model receiving the first prompt ([Fig. 15, 1506], [A request for “updating the amount of opportunity” indicates the request to be a prompt]), a third prompt for a user including a second query to clarify at least one of the first query and first data ([0210] The conversational chat assistant asks the user to clarify the record to update at 1508 by generating novel text via a generative language model, [In view of the generated third prompt 1508 of Fig. 15 which is clearly for purposes of clarifying the first query 1506 introducing opportunity, i.e. first content, through an additional question to target which opportunity]);
transmit, to the user interface ([0210] user interface 1500), the third prompt ([Fig. 15, 1508]);
obtain, via the user interface, a response to the third prompt ([Fig. 15, 1510]), the response to the third prompt including the second data clarifying at least one of the first query and the first data ([In view of the conversational history of Fig. 15 wherein “Acme” is second data clarifying the first query and/or data, i.e. updating opportunity]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 7, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the system of claim 1.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose: the one or more processors further configured to:
filter the first data from the first prompt according to a determination that the large language model is located at a first computing device distinct from a second computing device storing the first data,
wherein the first data is restricted from transmission according to a security policy that prevents transmission of the first data from the first computing device.
Kshirsagar discloses:
the one or more processors further configured to:
filter the first data from the first prompt according to a determination that the large language model is located at a first computing device distinct from a second computing device storing the first data ([Fig. 8, Mask Sensitive Information 810], [Fig. 9, User System 912 clearly distinct from Appl. Servers 950], [0316] retrieving information may involve accessing a data interface from retrieving information from another source, such as the Internet or a public or private data source residing outside of the database system, [0140] A determination is made at 810 as to whether to mask sensitive information…the determination may be made at least in part based on configuration information, [Explicitly disclosing a system architecture with a user system, i.e. first computing device, and another external source, i.e. second computing device, wherein the masking, i.e. filtering, determination is made based on configuration information including metadata, see [0186] (also including personal/private information, i.e. name/context/etc.), indicates that device location metadata, i.e. data retrieval location, will be included as configuration information and used for the masking determination in a system containing at least two distinct computing devices without extending beyond the scope of Kshirsagar]),
wherein the first data is restricted from transmission according to a security policy that prevents transmission of the first data from the first computing device ([0162] By way of example, different users may have different capabilities with regard to accessing and modifying application and database information, depending on a user's security or permission level, also called authorization. In systems with a hierarchical role model, users at one permission level may have access to applications, data, and database information accessible by a lower permission level user, but may not have access to certain applications, database information, and data accessible by a user at a higher permission level, [In view of the previously disclosed masking determination made based upon system architecture, it would not be unreasonable to state that a user’s security level is incorporated as configuration information used to determine when to mask restricted data, i.e. is the user authorized to mask/unmask personal information?, a form of database information modification]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 11, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
filtering the first data from the first prompt into the plurality of second prompts, each of the plurality of second prompts excluding the first data and including one or more second data clarifying the first query,
wherein the first data is restricted from transmission.
Kshirsagar discloses:
filtering the first data from the first prompt into the plurality of second prompts ([0141] Upon determining to mask sensitive information, sensitive information in the prompt is identified and replaced with unique identifiers at 812, [0148] Upon determining to replace sensitive information, the unique identifiers added to the prompt at 812 may be replaced with the corresponding sensitive information at 824, [Adding sensitive information back into a prompt after prompt completion, see Fig. 8, 816, indicates a filtering of the sensitive data from the first prompt, i.e. originally masked/incomplete, back into the second, i.e. complete, prompt]), each of the plurality of second prompts excluding the first data and including one or more second data clarifying the first query ([It is unclear to the examiner how the second prompts can possibly be excluding the first data when the first data has been previously filtered into the second prompts. Further, consider the unique identifiers associated with sensitive information to be first data associated with a first prompt/query after filtering and the replacement of these identifiers with the sensitive information to be second data clarifying the first query]),
wherein the first data is restricted from transmission ([Fig. 8, 822 Replace sensitive information?, No], [Opting to not replace identifiers with sensitive information indicates the sensitive information to be restricted from transmission]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 12, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
generating, by the large language model receiving the first prompt, a third prompt for a user including a second query to clarify at least one of the first query and first data;
transmitting, to the user interface, the third prompt;
obtaining, via the user interface, a response to the third prompt, the response to the third prompt including the second data clarifying at least one of the first query and the first data.
Kshirsagar discloses:
generating, by the large language model receiving the first prompt ([Fig. 15, 1506], [A request for “updating the amount of opportunity” indicates the request to be a prompt]), a third prompt for a user including a second query to clarify at least one of the first query and first data ([0210] The conversational chat assistant asks the user to clarify the record to update at 1508 by generating novel text via a generative language model, [In view of the generated third prompt 1508 of Fig. 15 which is clearly for purposes of clarifying the first query 1506 introducing opportunity, i.e. first content, through an additional question to target which opportunity]);
transmitting, to the user interface ([0210] user interface 1500), the third prompt ([Fig. 15, 1508]);
obtaining, via the user interface, a response to the third prompt ([Fig. 15, 1510]), the response to the third prompt including the second data clarifying at least one of the first query and the first data ([In view of the conversational history of Fig. 15 wherein “Acme” is second data clarifying the first query and/or data, i.e. updating opportunity]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 16, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the method of claim 10.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
filtering the first data from the first prompt according to a determination that the large language model is located at a first computing device distinct from a second computing device storing the first data,
wherein the first data is restricted from transmission according to a security policy that prevents transmission of the first data from the first computing device.
Kshirsagar discloses:
filtering the first data from the first prompt according to a determination that the large language model is located at a first computing device distinct from a second computing device storing the first data ([Fig. 8, Mask Sensitive Information 810], [Fig. 9, User System 912 clearly distinct from Appl. Servers 950], [0316] retrieving information may involve accessing a data interface from retrieving information from another source, such as the Internet or a public or private data source residing outside of the database system, [0140] A determination is made at 810 as to whether to mask sensitive information…the determination may be made at least in part based on configuration information, [Explicitly disclosing a system architecture with a user system, i.e. first computing device, and another external source, i.e. second computing device, wherein the masking, i.e. filtering, determination is made based on configuration information including metadata, see [0186] (also including personal/private information, i.e. name/context/etc.), indicates that device location metadata, i.e. data retrieval location, will be included as configuration information and used for the masking determination in a system containing at least two distinct computing devices without extending beyond the scope of Kshirsagar]),
wherein the first data is restricted from transmission according to a security policy that prevents transmission of the first data from the first computing device ([0162] By way of example, different users may have different capabilities with regard to accessing and modifying application and database information, depending on a user's security or permission level, also called authorization. In systems with a hierarchical role model, users at one permission level may have access to applications, data, and database information accessible by a lower permission level user, but may not have access to certain applications, database information, and data accessible by a user at a higher permission level, [In view of the previously disclosed masking determination made based upon system architecture, it would not be unreasonable to state that a user’s security level is incorporated as configuration information used to determine when to mask restricted data, i.e. is the user authorized to mask/unmask personal information?, a form of database information modification]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Regarding claim 20, Zhang in view of Chandrasekaran, further in view of Fabian discloses: the non-transitory computer readable medium of claim 19.
Zhang in view of Chandrasekaran, further in view of Fabian does not disclose:
the non-transitory computer readable medium further including one or more instructions executable by the processor to:
generate, by the large language model receiving the first prompt, a third prompt for a user including a second query to clarify at least one of the first query and first data;
transmit, to the user interface, the third prompt for a user including a second query to clarify at least one of the first query and the first data;
obtain, via the user interface, a response to the third prompt, the response to the third prompt including the second data clarifying at least one of the first query and the first data.
Kshirsagar discloses:
the non-transitory computer readable medium further including one or more instructions executable by the processor to:
generate, by the large language model receiving the first prompt ([Fig. 15, 1506], [A request for “updating the amount of opportunity” indicates the request to be a prompt]), a third prompt for a user including a second query to clarify at least one of the first query and first content ([0210] The conversational chat assistant asks the user to clarify the record to update at 1508 by generating novel text via a generative language model, [In view of the generated third prompt 1508 of Fig. 15 which is clearly for purposes of clarifying the first query 1506 introducing opportunity, i.e. first content, through an additional question to target which opportunity]);
transmit, to the user interface ([0210] user interface 1500), the third prompt for a user including a second query to clarify at least one of the first query and the first data ([Fig. 15, 1508], [Asking a user which opportunity to update is a clarifying second query]);
obtain, via the user interface, a response to the third prompt ([Fig. 15, 1510]), the response to the third prompt including the second data clarifying at least one of the first query and the first data ([In view of the conversational history of Fig. 15 wherein “Acme” is second data clarifying the first query and/or data, i.e. updating opportunity]).
Zhang, Chandrasekaran, Fabian, and Kshirsagar are considered analogous art within large language model prompt engineering/refinement. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Zhang in view of Chandrasekaran, further in view of Fabian to incorporate the teachings of Kshirsagar, because of the novel way to generate natural language text asking a user to clarify their intent with multiple iterations of feedback when ambiguity is identified in received user input, improving generated output in response to ambiguous questions (Kshirsagar, [0046]).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Zhang et al. (US-10225222-B2) discloses “Methods and apparatus directed to utilizing an automated messaging system to engage in a dialog with at least one user, via a computing device of the user, to determine a particular service entity for an action. In some implementations, the automated messaging system may generate a plurality of questions and/or other prompts to solicit user interface input from the user(s) for use in determining the particular service entity and/or in determining one or more criteria for the action. Some implementations are further directed to performing one or more computing actions based on the determined service entity and optionally based on one or more criteria for the action determined via user interface input of the dialog.” (abstract). See entire document.
Qazvinian et al. (US-20250013963-A1) discloses “The systems and methods described herein provide intelligent people analytics from generative artificial intelligence. In one embodiment, the system: receives a prompt related to people analytics from a client device associated with a user; generates an embedding representation of the received prompt using a generative AI system including one or more generative AI models; performs a similarity search using the generated embedding representation to identify similar prompts that have been submitted before; obtains an executable expression for responding to the received prompt; executes the executable expression using a data warehouse comprising one or more data sources to obtain a response to the received prompt; determines a type of response based on the nature of the received prompt; generates a response output based on the determined type and the response to the received prompt; and provides the response output to the client device associated with the user.” (abstract). See entire document.
Muthu et al. (US-20250209372-A1) discloses “Aspects of the present disclosure relate to generating optimized machine learning model prompts. Embodiments include providing an input prompt to a child machine learning model that directs the child machine learning model to generate an output. Embodiments further include generating a parent model prompt comprising instructions to generate a score for the input prompt based on one or more scoring criteria, the input prompt, and the output of the child machine learning model. Embodiments further include providing the parent model prompt to a parent machine learning model. Embodiments further include generating, by a generative machine learning model, an optimized prompt for the child machine learning model based on the generated score for the input prompt.” (abstract). See entire document.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to THEODORE JOHN WITHEY whose telephone number is (703)756-1754. The examiner can normally be reached Monday - Friday, 8am-5pm.
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, Andrew Flanders can be reached at (571) 272-7516. 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.
/THEODORE WITHEY/Examiner, Art Unit 2655
/ANDREW C FLANDERS/Supervisory Patent Examiner, Art Unit 2655
Prosecution Insights