DETAILED ACTION
Status of Claims
The present application, filed on or after 3/16/2013, is being examined under the first inventor to file provisions of the AIA .
This action is in reply to the Remarks and Amendments filed 08/03/2026.
Claims 1-20, i.e. all claims, have been amended.
Claims 1-20 have been examined and are pending.
(AIA ) Examiner Note
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 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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned at the time any inventions covered therein were effectively filed absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned at the time a later invention was effectively filed in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Priority
Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, or 365(c) is acknowledged.
However, Applicant has not complied with one or more conditions for receiving the benefit of an earlier filing date under 35 U.S.C. 120 as follows: The provisional application (63/658,788 filed 06/11/2024) does not contemplate the features of claims 4, 5, 11, 12, 18, 19 directed towards:
[claims 4, 11, 18]: “…an LLM instruction to generate the insight summaries with a recitation of an identifier of a location in the portion of the feedback data corresponding to the generated insight summary to reduce LLM hallucinations.”
[claims 5, 12, 19]: “…an LLM instruction to generate the primary insight summaries with a recitation of the insight summaries corresponding to and used to generate each primary insight summary to reduce LLM hallucinations.”
The later-filed application must be an application for a patent for an invention which is also disclosed in the prior application (the parent or original non-provisional application or provisional application). The disclosure of the invention in the parent application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994). Therefore, the earliest filing date afforded the claimed invention is that of the filing date of the presently filed claims (06/11/2025).
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea (i.e. a judicial exception) without significantly more.
Per step 1 of the Subject Matter Eligibility Guidance outlined in the MPEP 2106, the claims are directed towards a process, machine, or manufacture.
Per step 2A Prong One, the claims recite specific limitations which fall within at least one of the groupings of abstract ideas enumerated in the MPEP 2106, as follows:
Per Independent claims 1, 8, 15:
selecting a portion of the feedback data;
issuing an insight extraction prompt to the LLM, the insight extraction prompt comprising LLM instructions to determine whether the selected portion of the feedback data, contextually matches the insight type, the insight extraction prompt further comprising LLM instructions to generate an insight summary of the matched insight type from the selected portion of the feedback data;
repeating the issuing the insight extraction prompt for additional selections of the feedback data, thereby gathering additional insight summaries of the insight type;
issuing a grouping prompt to the LLM, the grouping prompt comprising LLM instructions to group similar insight summaries, and to generate a primary insight summary for each group, wherein the insight extraction prompt and the grouping prompt comprise LLM instructions to generate an LLM-stated source of an output of the LLM;
detecting LLM hallucinations, by comparing the output of the LLM with the LLM-stated source, relative to a prompt under which the LLM generated the output;
reducing the LLM hallucinations for subsequent LLM prompts based on using the LLM stated source when issuing the insight extraction prompt to the LLM and the grouping prompt to the LLM;
generating … a depiction of a chain of intermediate insight summaries from a primary insight summary to portions of the feedback data, the chain corresponding to relationships being used to generate the primary insight summaries.
As noted supra, these limitations fall within at least one of the groupings of abstract ideas enumerated in MPEP 2106. Specifically, these limitations fall within the group Certain Methods Of Organizing Human Activity (e.g. fundamental economic principles or practices (including hedging, insurance, mitigating risk); commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations); managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions).
That is, the steps, as drafted, are all directed towards a business decision to provide generic instructions to a pre-existing LLM (the LLM itself not being applicant’s invention) with the intention of summarizing data (i.e. customer feedback data) where such summarization includes a chain of citations of reference from where the data being summarized may be found and thus falling into Certain Methods of Organizing Human Activity; e.g. receiving/listening to customer feedback, then observing and honestly analyzing the customer feedback to find trends or correlations which align with pre-defined hypothesis (applicant’s “insight types”) with citations (to help ensure the summary is anchored in reality and truth) is nothing more than normal business practice and are activities routinely performed by sales and marketing managers in businesses who often instruct subordinates such as market analysts (analogous to the generic LLM being used per the claims) to provide such summaries and include their citations to references within market data (which includes customer feedback data). The use of readily available AI tools such as LLMs in a manner in which such LLMs are intended to be used (e.g. instructing and prompting them to interact with data and teaching them what type of analysis is desired) is not significantly more than this abstract idea.
For example, there is no technical improvement to the LLMs themselves. Instead, the applicant’s entire invention is directed towards manners of instructing and prompting and interacting with pre-existing LLM model(s) for the purpose of performing the aforementioned fundamental business sales and marketing practices. Such interaction and/or teaching with pre-established rules are all abstract ideas per Alice, Bilski, and MPEP 2106 and thus falling into Certain Methods Of Organizing Human Activity. Furthermore, the mere nominal recitation of a generic computer components used to facilitate the abstract idea does not take the claim limitation out of the enumerated grouping. Thus, the claims recite an abstract idea.
Per step 2A Prong 2, the Examiner finds that the judicial exception is not integrated into a practical application. Although there are additional elements, other than those noted supra, recited in the claims, none of these additional element(s) or a combination of elements as recited in the claims apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that it is more than a drafting effort designed to monopolize the exception. As drafted, the claims as a whole merely describe how to generally “apply” the aforementioned concepts (within a generic LLM framework) or, link them to a field of use (i.e. use of existing LLMs to summarize customer feedback datasets) or, serve as insignificant extra-solution activity (e.g. mere data-gathering, transfer, storage, and retrieval, etc…). The claimed computer components are recited at a high level of generality and are merely invoked as tools to implement the idea but are not technical in nature. Simply implementing the abstract idea on or with generic computer components is not a practical application of the abstract idea.
These additional limitations are as follows: “receiving a list of insight types; receiving feedback data;… for an insight type, inputting the selected portion of the feedback data to a large language model (LLM); …a user interface element…; wherein the LLM comprises at least one artificial intelligence (Al) model, executable on one or more processors, coupled to a memory, programmed with executable instructions, to store in the memory, a plurality of neurons organized in an array, wherein each neuron comprises a register, a processing element, and at least one neuron input, and a plurality of synaptic circuits, each synaptic circuit including a synaptic memory for storing a synaptic weight, wherein each neuron is connected to at least one other neuron via one of the plurality of synaptic circuits, wherein the AI model adjusts the synaptic weights, in a plurality of executions or runs of the AI model to generate an output.”
However, these elements do not present a technical solution to a technical problem; i.e. Applicant’s invention is not a technique nor technical solution for “receiving” data regardless of the description of the data (e.g. customer response data, called “feedback data”; and categories of data relationships called “insight types”) nor is it a technique nor technical solution for “inputting” such data into generic LLM model(s); i.e. LLMs necessarily operate on data and require some mechanism for “input”. Such inputting is so generic and recited at a high-level of generality that it includes general data-transfer and/or storage and/or manual entry via GUI, etc… Additionally, generation of user interface elements are akin to displaying data which is insignificant extra-solution activity in regards to the identified abstract idea and is not significantly more than the abstract idea itself. Furthermore, the “wherein” clause describing the generic LLM is indeed completely generic. At the time of filing, the applicant’s description of the LLM was already notoriously well-known and at such a high-level of generality encompasses the publicly available LLM model known as ChatGPT (e.g. Open AI’s Chat GPT-4 which is understood to be built upon artificial neural networks, specifically a Transformer architecture, that is inspired by, and designed to mathematically mimic, the structure of human neural ‘synaptic circuits’ which are weighted and requires memory to store such weightings. Furthermore, modern OpenAI models use Grouped-Query Attention (GQA), e.g. as opposed to standard multi-head attention, to improve efficiency. This means multiple query heads share fewer key-value heads etc…). This description is at a very high-level of generality and only describes the most basic of principles of the structure of known very complex LLM models. The additional elements do not recite a specific manner of performing any of the steps core to the already identified abstract idea. Instead, these features merely serve to generally “apply” the aforementioned concepts or, link them to a field of use (e.g. modeling using existing LLM models) or, are insignificant extra-solution activity (e.g data-gathering, input, storage, transfer, etc…) as pertains to the already identified abstract idea and do not integrate the abstract idea into a practical application thereof.
Per Step 2B, the Examiner does not find that the claims provide an inventive concept, i.e., the claims do not recite additional element(s) or a combination of elements that amount to significantly more than the judicial exception recited in the claim. As discussed with respect to Step 2A Prong Two, the additional elements in the independent claims were considered as merely serving to generally “apply” the aforementioned concepts via generically described computer components (e.g. non-transitory CRM and/or system comprising one or more processors, wherein the one or more processors are configured to perform [the claimed method steps]) and “link” them to a field of use (i.e. use of existing LLMs to summarize datasets), or as insignificant extra-solution activity. For the same reason these elements are not sufficient to provide an inventive concept; i.e. the same analysis applies here in 2B. Mere instructions to apply an exception using a generic computer component and conventional data gathering cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B. So, upon revaluating here in step 2B, these elements are determined to amount to no more than mere instructions to apply the exception using generic computer components (i.e. a server) and/or gather and transmit data which is well-understood, routine, conventional activity in the field; i.e. note the Symantec, TLI, and OIP Techs Court decisions cited in MPEP 2106.05(d)(ll) indicate that mere receipt or transmission of data over a network is a well-understood, routine, and conventional function when it is claimed in a merely generic manner (as it is here).
Accordingly, alone and in combination, these elements do not integrate the abstract idea into a practical application, as found supra, nor provide an inventive concept, and thus the claims are not patent eligible.
As for the dependent claims, the dependent claims do recite a combination of additional elements. However, these claims as a whole, considered either independently or in combination with the parent claims, do not integrate the identified abstract idea into a practical application thereof nor do they provide an inventive concept.
For example, dependent claims 4, 11, 18 recite the following: “…prompt further comprises an LLM instruction to generate the insight summaries with a recitation of an identifier of a location in the portion of the feedback data corresponding to the generated insight summary to reduce hallucinations.” Respectfully, this is nothing more than an explicit request (i.e. the prompt) to include citations to data (an identifier of a location in the feedback data) supporting the summary which is simply a good data-analytics practice (i.e. grounding the analysis in truth and facts is necessary for an honest analysis and summary of data which is for reducing error – applicant’s hallucinations). Therefore, this is part of the abstract idea as it falls squarely within routine marketing and sales analysis practices and therefore is a Method of Organizing Human Activity.
Therefore, the Examiner does not find that these additional claim limitations integrate the abstract idea into a practical application nor provide an inventive concept. Instead, these limitations, as a whole and in combination with the already recited claim elements of the parent claims, are not significantly more than the already identified abstract idea. A similar finding is found for the remaining dependent claims.
For these reasons, the claims are not found to include additional elements that are sufficient to amount to significantly more than the judicial exception and therefore the claims are not found to be patent eligible.
Please see the MPEP 2106 and the 2019 Revised Patent Subject Matter Eligibility Guidance published in the Federal Register (84 FR 50) on January 7, 2019 (found at http://www.uspto.gov/patent/laws-and-regulations/examination-policy/examination-guidance-and-training-materials).
Claim Rejections - 35 USC § 103 (AIA )
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 of this title, 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.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
Claims 1, 4, 5, 8, 11, 12, 15, 18, 19 are rejected under 35 U.S.C. 103 as obvious over Aggarwal et al. (U.S. 2025/0094732 A1; hereinafter, "Aggarwal") in view of Russell et al. (US 20240296295 A1; hereinafter, "Russell "), further in view of NPL (https://web.archive.org/web/20230424213457/ https://medium.com/@fenjiro/chatgpt-gpt-4-how-it-works-10b33fb3f12b; 4/17/2023).
Claims 1, 8, 15 (Currently amended)
Pertaining to claims 1, 8, 15 exemplified in the limitations of method claim 1, Aggarwal as shown teaches the following:
A computer implemented method comprising:
receiving a list of insight types1 (Aggarwal, see at least [0008], e.g. receiving “prioritized entity relationship categories” [insight types]; “…the prioritized entity relationship categories [insight types] identified in [received from] reference information stored by the system.…”);
receiving feedback data (Aggarwal, see citations noted supra, including again [0008] and [0034], regarding use of “…content [feedback data] to be summarized…”; a portion of this content, i.e. ‘the content to be summarized’, is disclosed as analyzed by an LLM. Although Aggarwal may not explicitly state he “receives” his “content” [feedback data], the Examiner nonetheless finds that by the very nature of the techniques Aggarwal is using, per his disclosure, some mechanism is necessary by which to receive such content and input it to his LLM to enable his use thereof as disclosed. Therefore, there is motivation to receive his “content” [feedback data] and it would therefore have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have explicitly received his content, to enable its use as disclosed, because per MPEP 2143(I) (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference teachings to arrive at the claimed invention is obvious. The motivation may be implicit and may be found in the knowledge of one of ordinary skill in the art, or, in some cases, from the nature of the problem to be solved. Id. at 1366, 80 USPQ2d at 1649.);
selecting a portion of the feedback data (Aggarwal, see citations noted supra, e.g. [0008]: “…content to be summarized…”; i.e. the portion of content [feedback data] which is to be summarized is a ‘portion of the content’ [portion of the feedback data]; see also [0041] and [0047]);
for an insight type, inputting the selected portion of the feedback data to a large language model (LLM); issuing an insight extraction prompt to the LLM, the insight extraction prompt comprising LLM instructions to determine whether the selected portion of the feedback data contextually matches the insight type (Aggarwal, see citations noted supra, further in view of Fig. 4 and [0053], teaching: “…The prioritized entity relationships [insights] correspond to one or more prioritized entity relationship categories [insight types]…” and “…the entities and the entity relationships [insights] for the content to be summarized [selected portion of feedback data] are identified by the subsystem 118 using a model 314 (e.g., an LLM) [i.e. inputting the selected portion of the feedback data to a large language model (LLM)] and an entity-relationship [insight] extraction prompt 302 [an insight extraction prompt]. The entity relationship extraction prompt 302 [insight extraction prompt] represents a small number of examples (e.g., 1-5) [instructions] that are used by the LLM to identify entities and entity relationships from the content to be summarized. The identified entities and entity relationships are then categorized using entity and entity relationship categories [insight types] information 312. Additional details of the categorization of entities and entity relationships is described in detail in FIG. 4 below. …”; i.e. the issued prompt is a set of instructions given to the LLM to determine whether the content to be summarized [selected portion of feedback data] contextually matches given entity relationship categories [insight types]; see also example provided per [0056]-[0070].),
the insight extraction prompt further comprising LLM instructions to generate an insight summary of the matched insight type from the selected portion of the feedback data (Aggarwal, may not explicitly teach that his “entity relationship extraction prompt 302” [insight extraction prompt] includes an explicit prompt to summarize the data which matches entity relationship categories [insight types]. However, Aggarwal per citations noted supra, e.g. per [0047] does teach a prompt to summarize such extracted entity relationship data, e.g. : “…The set of input parameters may include, for instance,… the prompts 114 for generating the multiple summaries [instructions to generate insight summary], the summarization strategy 106 (e.g., chain of thought prompting technique, a multi-hop question answer evaluation) used by the model to create the summaries, …, and so on…”; i.e. Aggarwal provides motivation to provide a prompt which comprises instructions to generate a summary of the extracted “entity relationships” and therefore it would have been obvious to include such prompt as part of his “entity relationship extraction prompt 302” [insight extraction prompt], e.g. for making his prompts more efficient and/or use fewer tokens – i.e. reduce the cost of using/prompting such LLMs, etc… and because per MPEP 2143(I) (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference teachings to arrive at the claimed invention is obvious. The motivation may be implicit and may be found in the knowledge of one of ordinary skill in the art, or, in some cases, from the nature of the problem to be solved. Id. at 1366, 80 USPQ2d at 1649.);
repeating the issuing the insight extraction prompt for additional selections of the feedback data, thereby gathering additional insight summaries of the insight type (Aggarwal, see citations noted supra, including again [0041], teaching e.g.: “… the prompts 114 for generating the multiple summaries…”; i.e. multiple prompts exist and as previously shown one prompt may be to extract entity relationships [insights] for a particular entity relationship category [insight type]. Furthermore, per [0092]-[0093], Aggarwal discloses that if no single summary is determined to be selected, e.g. no summary is selected as a “best” summary, etc…, then a responsive action may be taken such as “…For example, the responsive action taken by the subsystem can be to generate a new set of summaries [gathering additional insight summaries] for the content to be summarized and repeat the processing on the new set of summaries…”; Therefore, by extension, it appears that Aggarwal’s same prompt may be repeated but for additional content for the same “entity relationship category” [insight types] of interest. In view of these teachings, the Examiner finds there is motivation provided by Aggarwal to repeat the issuing of Aggarwal’s “entity relationship extraction prompt 302” [insight extraction prompt], for additional content [for additional selections of the feedback data] but for the same “entity relationship categories” [insight types] of interest, e.g. in an event when the initial content which was processed didn’t yield a best summary, e.g. perhaps when the initial content only sparsely referenced the desired entity relationship category [insight type], and therefore more data should be considered, etc… Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have performed a step of repeating the issuing of Aggarwal’s “entity relationship extraction prompt 302” [insight extraction prompt], for additional content [selections of the feedback data] for the same “entity relationship categories [insight types]” thereby generating a new set of summaries [gathering additional insight summaries] of the “entity relationship categories” [insight types] of interest);
issuing a grouping prompt to the LLM, the grouping prompt comprising LLM instructions to group similar insight summaries, and to generate a primary insight summary for each group (Aggarwal, see citations noted supra, including also at least Fig. 5, [0054], and [0081]-[0085], teaching: “FIG. 5 depicts a simplified flowchart 500 depicting the processing performed by subsystem 118 for generating clusters of summaries [group similar insight summaries] …At block 508, the subsystem uses a clustering technique to cluster the multiple summaries based on the incidence vectors generated for each summary. In a certain implementation, the subsystem uses a K-means clustering technique to generate one or more clusters of summaries where each cluster comprises one or more incidence vectors that represent one or more summaries from the multiple summaries. The K-means clustering technique is a commonly used clustering technique that is used to cluster data points into distinct groups (i.e., into a predefined number of k-clusters)…” and “…FIG. 5. The subsystem 118 then selects a cluster of summaries for further analysis using the incidence vectors and entity and entity relationship information extracted for each summary 310. The subsystem 118 analyzes the summaries within a selected cluster to identify [generate] a particular summary [a primary insight summary] that is deemed to be the "best" among the multiple generated summaries for the content to be summarized…”),
[…]
wherein the LLM comprises at least one artificial intelligence (AI) model, executable on one or more processors, coupled to a memory, […] (Aggarwal, see citations noted supra, including again at least Fig. 1, 2, 10B, associated disclosure, etc… and [0041] and [0053], e.g.: “…For instance, in one example, the subsystem 118 may use a large multimodal model (e.g., GPT-4) [LLM comprises at least one AI model executable on one or more processors] to identify entities and entity relationships from the content to be summarized, without specific training examples. In other examples, the subsystem 118 may use a pre-trained model (e.g., a BERT model) that uses named-entity recognition and entity linking for extracting entities and entity relationships from the content to be summarized…”; again per [0010] “…Various embodiments are described herein, including methods, systems, non-transitory computer-readable storage media storing programs, code, or instructions executable by one or more processors, and the like…”)
Although Aggarwal teaches the above limitations, and as noted supra Aggarwal teaches an extraction prompt and also a grouping prompt which generate a “best” summary [primary insight summary] from a group of summaries of content, Aggarwal may not explicitly delve into the below nuances directed towards prompting the LLM for source information which was used to generate such output (i.e. summaries and best summary). However, regarding the feature as recited below, Aggarwal in view of Russell teaches the following:
wherein the insight extraction prompt and the grouping prompt comprises LLM instructions to generate an LLM-stated source of an output of the LLM (Russel, see at least Abstract, [0002]-[0003], [0012], and [0026]-[0030], e.g.: “…Systems and methods for verifying attribution of quotations, generated by a large language model (LLM), to a source document are disclosed herein. Upon a request to summarize a source document or process a question that is answerable from a document, an LLM prompt is formed with the request or question along with the content of the source document. The LLM prompt is configured to cause an LLM to generate quotes that are intended to be from the source document. The output of the LLM, including the quotes, is then verified against the source document…”);
detecting LLM hallucinations, by comparing the output of the LLM with the LLM-stated
source, relative to a prompt under which the LLM generated the output (Russel, see citations noted supra, e.g. again per at least [0002]: “…The present technology provides, among other things, programmatic solutions to reducing the likelihood that a large language model (LLM) will return hallucinated content and/or verifying that the responses produced by the LLM are actually supported by the document for which the request was generated. For instance, an LLM prompt for requesting the summarization of a source document and/or providing a question about the source document is generated to encourage the reduction of hallucinated content from the LLM…”)
reducing the LLM hallucinations for subsequent LLM prompts based on using the LLM stated source (Russel, see citations noted supra, e.g. again per Abstract: “…The LLM prompt is configured to cause an LLM to generate quotes that are intended to be from the source document. The output of the LLM, including the quotes, is then verified against the source document…” and per [0002]: “…The present technology provides, among other things, programmatic solutions to reducing the likelihood that a large language model (LLM) will return hallucinated content and/or verifying that the responses produced by the LLM are actually supported by the document for which the request was generated.,,” and per at least [0026]: “…LLM prompt to produce an answer to a question about the document. The LLM prompt is generated to cause the output of the LLM prompt to include verbatim quotes from the source document 222 and/or source identifiers for the statements or quotes within the output.…”; verification against a source document is for the purpose of reducing “hallucinations” which are quotes which cannot be verified and/or which are fabricated by the LLM; see also again at least [0029]: “…The <Instructions> provide specific elements that encourage the LLM to produce accurate, verifiable results. For instance, the instructions clarify that answers are only to be drawn from the document that is included in the prompt. The instructions further clarify that the answer should be in the form of verbatim text. The format of the output is also defined within the instructions…”) when issuing the insight extraction prompt to the LLM and the grouping prompt to the LLM (Aggarwal already teaches issuing insight extraction prompts and grouping prompts as noted supra and it would therefore be obvious to implement Russel’s verification of citations and quotes against source documents, in response to, i.e. when issuing subsequent such prompts because per MPEP 2143(I) (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention is obvious. The motivation to combine may be implicit and may be found in the knowledge of one of ordinary skill in the art, or, in some cases, from the nature of the problem to be solved. Id. at 1366, 80 USPQ2d at 1649. );
generating a user interface element comprising a depiction of a chain of intermediate
insight summaries from a primary insight summary to portions of the feedback data, the chain corresponding to relationships being used to generate the primary insight summaries (Russel, see citations noted supra, in view of at least Fig. 3, [0008] and [0045]-[0051], e.g. per [0008]: “…FIG. 3 depicts an example application user interface for a productivity application displaying a source document and a verified LLM output….” And per [0018]: “…For example, an application UI 106 may be presented on the display screen 104, and the application UI 106 may display the content of a source document and a sidebar or interface for requesting information about the document. In some examples, the operating environment is a multi-application environment by which a user may view and interact with multiple productivity applications 110 through multiple application Uis 106…”)
Therefore, the Examiner understands that the limitation in question is merely applying known techniques of Russell (directed towards techniques of prompting LLMs for source information such as content of the source document [instructions to record a source of an output of the LLM] when asking the LLM to summarize a document and also detecting hallucination by verifying output of the LLM against source information based on the prompt to the LLM for a summary of content, etc…) which are applicable to a known base device/method of Aggarwal (already directed towards systems/method by which to prompt LLMs for a best summary [primary insight summary] of content which would benefit from source verification techniques) to yield predictable results. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the technique of Russell to the device/method of Aggarwal in order to perform the limitations in question because Aggarwal and Russell are analogous art in the same field of endeavor (at least G06Q30/02) and because according to MPEP 2143(I) (C) and/or (D), the use of known technique to improve a known device, methods, or products in the same way (or which is ready for improvement) is obvious.
Although Aggarwal/Russell teaches the above limitations including his LLM may be GPT-4, a reference to OpenAI’s ChatGPT-4, or other such LLM models, he may not explicitly delve into explaining the general minutia of the structure of GPT-4 and other such models as recited below. However, Aggarwal in view of NPL teaches these features, as follows:
[wherein the LLM…] programmed with executable instructions, to store in the memory, a plurality of neurons organized in an array, wherein each neuron comprises a register, a processing element, and at least one neuron input, and a plurality of synaptic circuits, each synaptic circuit including a synaptic memory for storing a synaptic weight, wherein each neuron is connected to at least one other neuron via one of the plurality of synaptic circuits, wherein the Al model adjusts the synaptic weights, in a plurality of executions or runs of the Al model to generate an output (ANPL, e.g. [pgs. 3-6], teaches OpenAI’s GPT-4 model comprises the following features: an artificial intelligence (AI) model; executable on one or more processors; necessarily coupled to a memory; is programmed and has functionality [i.e. with executable instructions], to store in memory a plurality of neurons organized in an array, wherein each neuron comprises a register, a processing element, and at least one neuron input, and a plurality of synaptic circuits, each synaptic circuit including a synaptic memory for storing a synaptic weight, wherein each neuron is connected to at least one other neuron via one of the plurality of synaptic circuits, wherein the Al model adjusts the synaptic weights, in a plurality of executions or runs of the Al model to generate an output; e.g.:
PNG
media_image1.png
752
808
media_image1.png
Greyscale
In view of these findings, the Examiner understands that Aggarwal provides motivation to implement an LLM, such as Chat GPT-4, which has the functionality and structure as disclosed by NPL. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date to have implemented such an LLM because per MPEP 2143(I) (B) Simple substitution of one known element for another to obtain predictable results is obvious and/or because per MPEP 2143(I) (G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention is obvious. The motivation to combine may be implicit and may be found in the knowledge of one of ordinary skill in the art, or, in some cases, from the nature of the problem to be solved. Id. at 1366, 80 USPQ2d at 1649.)
Claims 4, 11, 18 (Currently amended)
Although Aggarwal/NPL teaches the limitations upon which these claims depend, Aggarwal may not explicitly delve into the below nuances. However, regarding the feature as recited below, Aggarwal in view of Russell teaches the following:
…wherein the insight extraction prompt further comprises an LLM instruction to generate the insight summaries with a recitation of an identifier of a location in the portion of the feedback data corresponding to the generated insight summary to reduce hallucinations (Russel, see at least [0002]-[0003] and [0012], e.g.: “…Systems and methods for verifying attribution of quotations, generated by a large language model (LLM), to a source document are disclosed herein. Upon a request to summarize a source document or process a question that is answerable from a document, an LLM prompt is formed with the request or question along with the content of the source document. The LLM prompt is configured to cause an LLM to generate quotes that are intended to be from the source document. The output of the LLM, including the quotes, is then verified against the source document…”).
Therefore, the Examiner understands that the limitation in question is merely applying a known technique of Russell (directed towards techniques of prompting LLMs for source information such as content of the source document [a recitation of an identifier of a location in the portion of the feedback data corresponding to the generated insight summary] when asking the LLM to summarize a document) which is applicable to a known base device/method of Aggarwal (already directed towards systems/method by which to prompt LLMs for summaries of content) to yield predictable results. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the technique of Russell to the device/method of Aggarwal in order to perform the limitation in question because Aggarwal and Russell are analogous art in the same field of endeavor (at least G06Q30/02) and because according to MPEP 2143(I) (C) and/or (D), the use of known technique to improve a known device, methods, or products in the same way (or which is ready for improvement) is obvious.
Claims 5, 12, 19 (currently amended)
Although Aggarwal/NPL teaches the limitations upon which these claims depend, and Aggarwal teaches a grouping prompt to generate a “best” summary [primary insight summary] from a group of summaries of content, Aggarwal may not explicitly delve into the below nuances directed towards prompting the LLM for source information used to generate such summary. However, regarding the feature as recited below, Aggarwal in view of Russell teaches the following:
…wherein the grouping prompt further comprises an LLM instruction to generate the primary insight summaries with a recitation of the insight summaries corresponding to and used to generate each primary insight summary to reduce hallucinations (Russel, see at least [0002]-[0003] and [0012], e.g.: “…Systems and methods for verifying attribution of quotations, generated by a large language model (LLM), to a source document are disclosed herein. Upon a request to summarize a source document or process a question that is answerable from a document, an LLM prompt is formed with the request or question along with the content of the source document. The LLM prompt is configured to cause an LLM to generate quotes that are intended to be from the source document. The output of the LLM, including the quotes, is then verified against the source document…”).
Therefore, the Examiner understands that the limitation in question is merely applying a known technique of Russell (directed towards techniques of prompting LLMs for source information such as content of the source document [a recitation of an identifier of a location in the portion of the feedback data corresponding to the generated insight summary] when asking the LLM to summarize a document) which is applicable to a known base device/method of Aggarwal (already directed towards systems/method by which to prompt LLMs for a best summary [primary insight summary] of content) to yield predictable results. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the technique of Russell to the device/method of Aggarwal in order to perform the limitation in question because Aggarwal and Russell are analogous art in the same field of endeavor (at least G06Q30/02) and because according to MPEP 2143(I) (C) and/or (D), the use of known technique to improve a known device, methods, or products in the same way (or which is ready for improvement) is obvious.
Claims 2, 9, 16 are rejected under 35 U.S.C. 103 as obvious over Aggarwal, Russel, and NPL further in view of Liu et al. (U.S. 20250181224 A1; hereinafter, "Liu")
Claims 2, 9, 16
Although Aggarwal/Russel/NPL teaches the limitations upon which these claims depend, and Aggarwal teaches grouping prompts as noted supra, e.g. per Fig. 5, [0054], and [0081]-[0085], teaching: “FIG. 5 depicts a simplified flowchart 500 depicting the processing performed by subsystem 118 for generating clusters of summaries [group similar insight summaries] …At block 508, the subsystem uses a clustering technique to cluster the multiple summaries …”, he may not explicitly delve into a technique by which to accommodate input threshold limits established for various LLMs nor specifically teach batching such input based on such thresholds. However, regarding these features Aggarwal/Russell in view of Liu teaches the following:
…further comprising:
when a number of insight summaries of a type exceeds a selected LLM input size threshold, performing preliminary grouping by: batching the insight summaries into batches smaller in size than the selected LLM input size threshold; processing each batch of insight summaries in the LLM with the grouping prompt, by: providing each batch of insight summaries to the LLM as input; (Liu, see at least [0025], e.g.: “…If the total size of the descriptions [e.g. insight summaries] is greater than an input limit [a selected input size threshold] of the LLM(s) [..they…] may be divided into a plurality of batches [performing preliminary grouping by batching the insight summaries into smaller sizes]. The descriptions [e.g. insight summaries] corresponding to each of the plurality of batches [similarly grouped summaries] may be sequentially input into the LLM(s) [providing the preliminary insight summaries to the LLM as input]. For example, descriptions [e.g. insight summaries] corresponding to the algorithm tools in a first batch [similarly grouped into a first batch] of the plurality of batches may be input into the LLM(s), then descriptions [e.g. insight summaries] corresponding to the algorithm tools in a second batch [similarly grouped into a second batch] of the plurality of batches may be input into the LLM(s), and so on...”)
and issuing the grouping prompt to the LLM, which instructs the LLM to group similar insight summaries of each batch, and to generate a primary insight summary for each group of similar insight summaries of each batch to reduce hallucinations (Aggarwal, see citations already noted supra, including also at least Fig. 5, [0054], and [0081]-[0085], teaching: “FIG. 5 depicts a simplified flowchart 500 depicting the processing performed by subsystem 118 for generating clusters of summaries [grouping similar insight summaries] …At block 508, the subsystem uses a clustering technique to cluster the multiple summaries based on the incidence vectors generated for each summary. In a certain implementation, the subsystem uses a K-means clustering technique to generate one or more clusters of summaries…” and “…FIG. 5. The subsystem 118 then selects a cluster of summaries for further analysis using the incidence vectors and entity and entity relationship information extracted for each summary 310. The subsystem 118 analyzes the summaries within a selected cluster to identify a particular summary [a primary insight summary] that is deemed to be the "best" among the multiple generated summaries for the content to be summarized…”)
Therefore, the Examiner understands that the limitations in question are merely applying known techniques of Liu (directed towards a technique of batching input data if the total size exceeds an input limit threshold of the LLM, etc…) which is applicable to a known base device/method of Aggarwal (already directed towards processing such input summary data through use of a grouping prompt, to group input provided to an LLM to generate a primary summary or “best” summary) to yield predictable results. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the technique of Liu to the device/method of Aggarwal in order to first check a pre-established threshold for input descriptions [e.g. insight summaries of a particular type] to an LLM and when a number of insight summaries of a type exceeds a selected LLM input size threshold, performing Liu’s technique of preliminary grouping by: batching such descriptions (e.g. the insight summaries) into batches smaller in size than the selected LLM input size threshold and then processing each batch of insight summaries in the LLM with the grouping prompt already taught and use per Aggarwal’s system/method because Aggarwal and Liu are analogous art in the same field of endeavor (at least G06F 40/40) and because according to MPEP 2143(I) (C) and/or (D), the use of known technique to improve a known device, methods, or products in the same way (or which is ready for improvement) is obvious.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(B) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3, 10, 17 are rejected under 35 U.S.C. 112(b) or (for pre-AIA ) 35 U.S.C. 112, second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor, a joint inventor, or (for pre-AIA ) the applicant regards as the invention.
Respectfully, dependent claims 3, 10, 17 have been amended to recited features which are unclear. For example, these claims recite in part the following:
PNG
media_image2.png
314
742
media_image2.png
Greyscale
Respectfully, the claim appears to contradict itself; i.e. the claim has a first feature which requires the following to occur: “…batching the insight summaries into batches smaller in size than the selected LLM input size threshold.” However, the claim then states the following: “…wherein when the preliminary insight summaries still exceed the selected LLM input size threshold: further batching…” Respectfully, either the first feature occurs as claimed and the second feature is extraneous and not needed nor implemented or, applicant does not have possession of a mechanism by which to fully implement the first feature, or applicant is implying something else, e.g. some step which is not recited but which is necessary causes some reversion of the first feature after it is performed such that the second feature is necessary; this later idea is indefinite for failure to include in the claims a necessary step. Regardless, because it is not clear what steps must occur, the claims are indefinite. A person of ordinary skill in the art would not be apprised of the full scope of conditions under which the claim is infringed and therefore the claims are indefinite.
There is a great deal of speculation necessary on the part of a person of ordinary skill in the art in interpreting the claim and therefore it would not be proper to provide an overly speculative interpretation. Correction or clarification is required.
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 7 and 14 are rejected under 35 U.S.C. 112(d), as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends.
Claims 7 and 14 now recite features which fail to further limit the method of claim 1 and/or the operations performed by the processing device executing instructions stored on the non-transitory CRM of claim 8. These features, comprising the entirety of the body of the claim, are a “wherein” clause which does not provide any action which must occur to infringe the claims; i.e. this clause is recited as follows: “…wherein a selected LLM input size threshold is a size selected to reduce LLM hallucination determined based a content window size parameter of the LLM, the content window size parameter of the LLM being expressed in units of tokens as a maximum number of tokens the LLM can analyze for one set of input data and for one prompt.”
Note, the parent claims (claims 1 and 8) do not recite any selection of an LLM input size threshold. Furthermore, no claim feature which depends from claims 7 and 14 further make use of, or depend on, any such selected LLM input size threshold. Furthermore, such a description, i.e. this “wherein” clause, as claimed herein, is not an action and is not a method step or operation. It is not clear what limitation applicant is implying is added to his parent claims or what a practitioner of his claimed invention must do or possess to infringe the invention as claimed per method claim 1 and/or the non-transitory CRM of claim 8. This wherein clause doesn't require the performance of anything; it is completely non-functional descriptive material and it is not clear what step must be performed as part of the method and/or operations being recited in the claims.
It is therefore not clear whether applicant is attempting to introduce a new and distinct limitation of some type or alter or change the scope of the already recited method or non-transitory CRM, or something else. It appears the feature fails to further limit the subject matter of the claims upon which it depends. Because of this confusion, the claim is not clear and held to be indefinite.
There is a great deal of speculation necessary on the part of a person of ordinary skill in the art in interpreting the claim and therefore it would not be proper to provide an overly speculative interpretation. Correction or clarification is required.
Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Examiner’s Note
Adhering to the guidelines of MPEP 2173.06(II), the Examiner acknowledges that all words in a claim must be considered in judging the patentability of a claim against the prior art. In re Wilson, 424 F.2d 1382, 165 USPQ 494 (CCPA 1970). The fact that terms may be indefinite does not make the claim obvious over the prior art. However, when the terms of a claim are considered to be indefinite, at least two approaches to the examination of an indefinite claim relative to the prior art are possible.
First, where the degree of uncertainty is not great, and where the claim is subject to more than one interpretation and at least one interpretation would render the claim unpatentable over the prior art, an appropriate course of action would be for the examiner to enter two rejections: (A) a rejection based on indefiniteness under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph; and (B) a rejection over the prior art based on the interpretation of the claims which renders the prior art applicable. See, e.g., Ex parte Ionescu, 222 USPQ 537 (Bd. App. 1984). When making a rejection over prior art in these circumstances, it is important for the examiner to point out how the claim is being interpreted.
Second, where there is a great deal of confusion and uncertainty as to the proper interpretation of the limitations of a claim, it would not be proper to reject such a claim on the basis of prior art. As stated in In re Steele, 305 F.2d 859, 134 USPQ 292 (CCPA 1962), a rejection under 35 U.S.C. 103 should not be based on considerable speculation about the meaning of terms employed in a claim or assumptions that must be made as to the scope of the claims.
In the present set of claims (filed8/3/2026), a great deal of confusion and uncertainty as to the proper interpretation of the limitations of dependent claims 3, 7, 10, 14, and 17 exists. Therefore, a rejection under 35 U.S.C. 103 is not made regarding these claims, because, as shown in the 35 USC § 112(b) rejections, speculation and conjecture must be utilized by the Office and the artisan inasmuch as the claims do not adequately reflect what the disclosed invention is under the second paragraph of 35 USC § 112. Again, see In re Steele, 305 F.2d 859, 862 (CCPA 1962) (A prior art rejection cannot be sustained if the hypothetical person of ordinary skill in the art would have to make speculative assumptions concerning the meaning of claim language.)
Response to Arguments
Applicant amended all claims, i.e. claims 1-20, on 08/03/2026. Applicant's arguments (hereinafter “Remarks”) also filed 08/03/2026, have been fully considered but are moot in view of the new grounds of rejection necessitated by applicant’s amendments. Note the new 35 USC 101, 112(b) and (d), and 103 rejections with Aggarwal in view of Russel, and NPL teaching applicant’s amended features of the independent claims.
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).
The following prior art is made of record although not relied upon as it is considered pertinent to applicant's disclosure:
US Publication US 20250156419 A1, Shah discusses a technical challenge is to incorporate the use of LLMs and/or other GAI models into the operational flows of application software systems while mitigating the risk of AI hallucination, etc… which appears to be a reference of importance and could be used as a substitute reference in the 35 USC 103 rejection provided supra.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHAEL J SITTNER whose telephone number is (571)270-3984. The examiner can normally be reached M-F; ~9:30-6:30. 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, Waseem Ashraf can be reached on (571) 270-3948. 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.
/Michael J Sittner/
Primary Examiner, Art Unit 3621
1 Spec appears to indicate applicant’s "insight types" means a category of data within a dataset, e.g. a category or type of customer "feedback" data (e.g. category of customer response data, such as within online customer reviews, survey results, customer service transcripts, etc.... ); e.g. per [0019] and [0045]: “…insight type, for
example "feature request," "complaint," and "customer pain point,"...”; i.e. data within customer feedback data may be categorized as a “complaint” or, as a product “feature request” or, as an issue with a product feature called a “pain point”, etc…