DETAILED ACTION
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 Amendment and Arguments.
Applicant’s arguments, see page 12, 13, and 14, filed 1/2/2026, with respect to claims 1-7, 9-13 rejection have been fully considered and are not persuasive. Applicant argues that “Applicant does not concede to these statements, or to the underlying 101 rejections. However, in order to expedite prosecution, Applicant presently amends the independent claims to recite, inter alia, "receiving, from a generative natural language model and based on a first textual prompt, a first output that represents a database entry" and "generating, via the generative natural language model, a second output based on the second textual prompt, wherein the second output represents an update to the database entry" (emphasis added). In view of the Examiner's suggestions in the AIIS, and the Applicant's clarifications above regarding the meaning of the claim term "database entry," Applicant submits that the claims, as presently amended, are directed to patent-eligible subject matter and requests that these rejections be withdrawn.” The Examiner respectfully disagrees adding generative to natural language model and database entry corresponding to a plurality of elements does not make the claim patent eligible. Adding generative to natural language model does not describe what the model is. Where a language model of a list of words (like a dictionary) which is used to generate a piece of information. Further, in regards to the database entry see paragraph 149 of the specification where it states “The embodiments described herein provide improvements with respect to these issues. These embodiments include obtaining a first user input (e.g., a name of a desired catalog object for which a database entry can be generated) from which can be generated a first textual prompt (e.g., by adding the text of the first user input to default text or other default prompt content specifying a format in which to generate outputs so as to facilitate parsing of the outputs into database entries). This first textual prompt is then applied to a natural language model (e.g., a large language model) or other generative model to generate a first output that represents a database entry. The model and/or first textual prompt are specified such that the database entry may include a plurality of elements. Such elements could include a long and/or short (e.g., summary) description of the database entry, elements corresponding to inputs to be obtained from a user in order to generate a request record according to the database entry (e.g., inputs that specify properties of a piece of hardware or software, requests for which can be generated according to the database entry), elements corresponding to steps that can be taken to fulfill a request generated according to the database entry, and/or some other elements.” This indicates that it could be a singular element (emphasis added). The claims are not patent eligible.
Therefore, the 101 rejection of claims 1-7, 9-13 are maintained.
Applicant’s amendment, for claims 14,15,17-18, 20-23 have been fully considered. The 101 rejection of claims 14,15,17-18, 20-23 has been withdrawn.
Applicant’s arguments with respect to prior art for claim(s) 8 have been considered but are not persuasive. Applicant argues in regard to 35 U.S.C 103 states that “However, this is an incorrect correspondence because these portions of Agastya merely generically teach "identif[ying] PII and other confidential text in [a] prompt." This does not amount to the amended claim 1 limitation that "a generative natural language model" generates, "based on a first textual prompt," an output that represents "a database entry" that includes both "a first element having a first data input" and also "an input sanitization criterion for the first data input." Indeed, even if we accepted the Office's correspondences as correct, Agastya would instead teach the "generative natural language model" "sanitizing" an input itself, rather than the claimed limitation that the "generative natural language model" outputs a representation of "an input sanitization criterion" that can be applied (e.g., by a computer determining "that the input be an integer" or evaluating "a regular expression that the input must satisfy" as provided in 0155 of Applicant's specification) to later-received "inputs."” The Examiner respectfully disagrees, the AI gateway tool would correspond to the generative language model since they both are participating in a prompt response workflow involving natural language. Aswell as, the AI gateway tool (“AI gateway tool which identifies PIT and other confidential information in a prompt”) teaches a database entry which contains (a first element, and input sanitization) since the claim doesn’t distinguish the two. claim 1 state ” first output that represents a database entry”, “wherein the database entry includes an input sanitization criterion for the first data input;” represents is a broad term which can include two outputs being a single output. Further, the claim doesn’t mention that’s its being applied or that it will be used at a later time. Hence the rejection is maintained.
In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., “a regular expression”) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Applicant’s arguments with respect to claim(s) 14, 15, 17, 18, 20-23 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 1-7, 9-13 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1 Further claim 1 recites A method comprising:
A method comprising: receiving, from a generative natural language model and based on a first textual prompt, a first output that represents a database entry, wherein the database entry includes a plurality of elements, wherein the plurality of elements includes a first element having a first data input, and wherein the database entry includes an input sanitization criterion for the first data input; obtaining a selection of a subset of the plurality of elements; determining a second textual prompt based on the selected subset of the plurality of elements; and generating, via the generative natural language model, a second output based on the second textual prompt, wherein the second output represents an update to the database entry.
The limitation of “receiving…”, “obtaining…”, “determining…”, and “generating…” , as drafted, is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. For example, a person receiving a query from a user and making sure that query follows a rule set. Further, answering it in a way where it contains a plurality of parts. Then, the user gives additional instruction by selecting a certain part from the plurality of parts. Further, the person adjusts his answer (in regards from the user’s further request) and gives an updated response for that part.
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 claims recite an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using the computer components amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claims are not patent eligible.
Claims 2 additionally recite the method of claim 1, further comprising updating the database entry according to the second output. However, this limitation does not prevent a human from performing the steps mentally as described above. Further, the person updating the response entirely and only adjusting the part that was further edited. Thus, these claims are directed towards a mental process. Similar to above, no additional limitations are provided that provide a practical application, or amount to significantly more than the abstract idea. Therefore, the claims are not patent eligible.
Claims 3 additionally recites the method of claim 1, wherein the second output represents an update to the selected subset of the plurality of elements of the database entry. However, these limitations encompass a person receiving a query from a user and answering that query in a plurality of parts. Then the user further selects a part and gives further modification for the person. Where the person redoes/edits that part while keeping the rest the same. Where the final output the user receives is all the parts (where the edited/modifies part replaces the select part the user)
Claims 4 additionally recites The method of claim 3, wherein obtaining the selection of the subset of the plurality of elements comprises receiving, from a user via a user interface, an input indicating the selected subset and receiving, from the user via the user interface, a selection of one modification option from an enumerated set of modification options, wherein the one modification option of the enumerated set of modification options is configured to receive free-form text from the user via the user interface. However, these limitations encompass a person receiving a query from a user and answering it in a way where it contains a plurality of parts. Then, the user gives additional instruction by selecting a certain part from the plurality of parts. Further, the person redoes/adjusts his answer (in regards from the user’s further request) and gives an updated response for that part while maintain the other parts. Thus, the claim is directed towards a mental process. In particular, the claim only recites additional elements that is “user interface” where its pre-solution by receiving something from the user. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
Claims 5 additionally recites The method of claim 3, wherein obtaining the selection of the subset of the plurality of elements comprises receiving, from a user via a user interface, an input indicating the selected subset of the plurality of elements and receiving, from the user via the user interface, a selection
PNG
media_image1.png
10
6
media_image1.png
Greyscale
of one modification option from an enumerated set of modification options, wherein the enumerated set of modification options includes options to elaborate, simplify, regenerate, or delete one of the elements of the selected subset or to create a new element within the selected subset. However, these limitations encompass a person receiving a query from a user and answering it in a way where it contains a plurality of parts. Then, the user gives additional instruction by selecting a certain part from the plurality of parts. Further, the person redoes/adjusts his answer (in regards from the user’s further request) and gives an updated response for that part while maintain the other parts. Thus, the claim is directed towards a mental process. In particular, the claim only recites additional elements that is “user interface” where its pre-solution by receiving something from the user. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
Claims 6 additionally recites the method of claim 1, wherein the generative natural language model has been trained using a plurality of additional database entries related to operation of a particular managed network. However, these limitations encompass a person being trained/and or having a specific field that he answers queries from a user. Thus, the claim is directed towards a mental process. Similar to above, no additional limitations are provided that provide a practical application, or amount to significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claims 7 additionally recites the method of claim 6, wherein at least a portion of the plurality of additional database entries were generated by the generative natural language model prior to being trained using the plurality of additional database entries. However, these limitations encompass a person answering query from a user and learning/remembering previous ones to answer future queries. Thus, the claim is directed towards a mental process. Similar to above, no additional limitations are provided that provide a practical application, or amount to significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claims 9 additionally recite The method of claim [[8]]1, wherein the method further comprises, prior to obtaining the selection of the subset of the plurality of elements: responsive to determining that the first element includes the first data input, determining a third textual prompt based on the first element; generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents the input sanitization criterion for the first data input; and updating the database entry according to the third output to include the input sanitization criterion for the first data input. However, these limitations encompass a person receiving a query from a user and making sure that query follows a rule set. Further, providing an answer that follows said ruleset like formatting. Thus, the claim is directed towards a mental process. Similar to above, no additional limitations are provided that provide a practical application, or amount to significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim 10 additionally recites The method of claim [[8]]1, further comprising: receiving, from a user via a user interface, a textual description of the input sanitization criterion; determining a third textual prompt based on the textual description, wherein the third textual prompt includes a request for a regular expression representing the input sanitization criterion as described by the textual description; generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output includes the regular expression that represents the input sanitization criterion; and updating the database entry according to the third output to include the regular expression as the input sanitization criterion for the first data input. However, these limitations encompass a person receiving from a user a rule set he has to follow. Then receiving a query where the person frequently receives and remembers how to answer it. Further providing that answer to the user. Thus, the claim is directed towards a mental process. In particular, the claim only recites additional elements that is “user interface” where its pre-solution by receiving something from the user. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
Claim 11 additionally recites the method of claim 10, further comprising: receiving, from the user via the user interface, a test input; determining whether the test input satisfies the regular expression; and providing, via the user interface, an indication of whether the test input satisfies the regular expression. However, these limitations encompass a person receiving from a user an input (test input) and the person remembers the answer from a previous query so he matches it and provides the answer. Further the user can determine whether the match was successful or not. Thus, the claim is directed towards a mental process. In particular, the claim only recites additional elements that is “user interface” where its pre-solution by receiving something from the user. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
Claim 12 additionally recites The method of claim 1, wherein the first output includes a listing of the elements of the plurality of elements, and wherein the method further comprises, for each element of the plurality of elements of the database entry: determining a third textual prompt based on the element, wherein the third textual prompt includes at least one of a request to re-generate the element, a request to generate an input sanitization criterion for the element, or a request to elaborate the element; generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents an update to the element; and updating the element of the database entry according to the third output. However, these limitations encompass , a person receiving a query from a user and answering it in a way where it contains a plurality of parts. Then, the user gives additional instruction by selecting a certain part from the plurality of parts. Further, the person adjusts/re-generates his answer (in regards from the user’s further request) and gives an updated response for that part while maintain the remaining parts
Claim 13 additionally recites The method of claim 1, wherein obtaining the selection of the subset of the plurality of elements includes receiving a command to re-generate the selected subset of the plurality of elements, and wherein determining the second textual prompt based on the selected subset of the plurality of elements comprises determining the second textual prompt to include a portion of the first output that represents the selected subset of the plurality of elements and a request to re- generate the elements represented by the portion of the first output. However, these limitations encompass , a person receiving a query from a user and answering it in a way where it contains a plurality of parts. Then, the user gives additional instruction by selecting a certain part from the plurality of parts. Further, the person adjusts/re-generates his answer (in regards from the user’s further request) and gives an updated response for that part while maintain the remaining parts
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.
The factual inquiries 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 nonobviousness.
Claims 1-7, 9 and 12-13 are rejected under 35 U.S.C. 102 (a)(2) as being anticipated by US Patent US 20250181824 A1, (Szabo; Jacint.), in view of US Patent US 20250131126 A1, (Kommanamanchi; Agastya.).
Claim 1
Regarding Claim 1 Jacint teach
receiving, from a generative natural language model and based on a first textual prompt, a first output that represents a database entry, wherein the database entry includes a plurality of elements, wherein the plurality of elements includes a first element having a first data input, and wherein the database entry includes an input sanitization criterion for the first data input;
(Fig 3A shows the user entering the prompt (element 372) and the plurality of elements below (element 354A)
FIG 3c see element 356B, Fig 3 D element 356C, Elements 354A, 356B, and 356C would be the plurality of elements. Also the plurality of elements can be a single element based on the specification (paragraph 146)
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
obtaining a selection of a subset of the plurality of elements;
(Fig 3B shows the user selecting an element (in 354A (356A)))
determining a second textual prompt based on the selected subset of the plurality of elements; and
(Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372))
generating, via the generative natural language model, a second output based on the second textual prompt, wherein the second output represents an update to the database entry.
(Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372)
Fig 3C shows the new output where only the selected portion was changed (element 356B)
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
Jacint do not explicitly teach all of wherein the database entry includes a plurality of elements, wherein the plurality of elements includes a first element having a first data input, and wherein the database entry includes an input sanitization criterion for the first data input;
However, Agastya teach wherein the database entry includes a plurality of elements, wherein the plurality of elements includes a first element having a first data input, and wherein the database entry includes an input sanitization criterion for the first data input;
(Paragraph 46-49 " Initially, at S210, a prompt 304 is received. The prompt 304 may be received from a user 105 in the “Message to be Vetted” user entry field 302. In FIG. 3, the prompt is “Does phone number 5555555555 belong to Joe.” In other embodiments, the prompt may be received automatically and directly from an application 103, without user entry of text in a field.
The user then selects the “submit” 310 icons in S211.
In S212, a personal identifiable information (PII) status (e.g., “Contains PII” or “PII-free”) is determined. PII uses data to confirm an individual's identity. Sensitive PII may include, but is not limited to, a full name, face, home address, social security number, passport number, birthdate, driver's license, financial information, medical records, finger prints or handwriting sample, email address, phone number, etc. The AI gateway tool 122 identifies PII and other confidential text in the prompt 304 via the image component 125 and the text component 127. As described above, the image component 125 and the text component 127 may analyze the prompt via an internal ML model to identify PII or confidential information, or may access an external service to identify PII so the image component 127 and text component 127 may determine the PII status (e.g., presence or absence of PII data). Continuing with the non-exhaustive example of FIG. 3, the text component 127 may determine the PII status of the prompt 304.
In a case it is determined at S212, PII status is “Contains PII”, the method proceeds to S214 and a “Contains PII” output 402 (FIG. 4) is returned to the display in the Vetted Response field 306. The output 402 includes the PII status, and in the case of a “Contains PII” status, the output 402 further describes the PII included in the prompt. Additionally, in the case of the “Contains PII” status, the “Add model” icon 314 and “Send to AI” icon 316 are greyed out (as shown in FIG. 4) and not selectable by the user in response to the PII status of “Contains PII”."
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Jacint to incorporate the teachings of Agastya to provide a “wherein the plurality of elements includes a first element having a first data input, and wherein the database entry includes an input sanitization criterion for the first data input.” Doing so would make system more susceptible to breaches and privacy violations , as recognized by Agastya. (Paragraph 34).
Claim 2
Regarding Claim 2, Jacint in view of Agastya , further Jacint teach
The method of claim 1, further comprising updating the database entry according to the second output.
(Fig 3c shows the new output where only the selected portion was changed (element 356B))
Claim 3
Regarding Claim 3, Jacint in view of Agastya , further Jacint teach
The method of claim 1, wherein the second output represents an update to the selected subset of the plurality of elements of the database entry.
(Fig 3c shows the new output where only the selected portion was changed (element 356B))
Claim 4
Regarding Claim 4, Jacint in view of Agastya , further Jacint teach
4. The method of claim 3, wherein obtaining the selection of the subset of the plurality of elements comprises receiving, from a user via a user interface,
(Figs 3A-3E shows examples of the User interface as well as selection of elements)
an input indicating the selected subset and receiving, from the user via the user interface,
(Fig 3B shows the user selecting an element (in 354A (356A)))
a selection of one modification option from an enumerated set of modification options, wherein the one modification option of the enumerated set of modification options is configured to receive free-form text from the user via the user interface.
(Fig 3B shows the user selecting an element (in 354A (356A)) as well as shows the user entering a second prompt (in text) where he wants it to start at a different time (element 372))
Claim 5
Regarding Claim 5, Jacint in view of Agastya , further Jacint teach
5. The method of claim 3, wherein obtaining the selection of the subset of the plurality of elements comprises receiving, from a user via a user interface,
(Figs 3A-3E shows examples of the User interface as well as selection of elements)
an input indicating the selected subset of the plurality of elements and receiving, from the user via the user interface,
(Figs 3A-3E shows examples of the User interface as well as selection of elements)
a selection of one modification option from an enumerated set of modification options, wherein the enumerated set of modification options includes options to elaborate, simplify, regenerate, or
(Fig 3D-3E (shows the user selecting a portion (element 356C) and wanting it in a paragraph form (element 372)) where in Fig 3E it shows a modified/regenerated portion ()element 356D)
delete one of the elements of the selected subset or to create a new element within the selected subset.
Claim 6
Regarding Claim 6, Jacint in view of Agastya , further Jacint teach the method of claim 1, wherein the generative natural language model has been trained using a plurality of additional database entries related to operation of a particular managed network.
(Paragraph 27 "Accordingly, in some implementations, the LLM may be trained and/or fine-tuned to process commands to account for discrepancies between facts or details contained inside and outside of a user's selection. For example, an explicit user follow up request to replace a first date with a second date in a selected portion of rendered LLM output (generated from an underlying LLM response) may trigger generation of an implied request to also replace the first date with the second date elsewhere in the rendered LLM output, even in portion(s) not selected by the user. This feature may be particularly useful when the LLM is used to generate complex structured language such as source code, mathematical proofs, etc. For instance, a user may select a particular code segment (e.g., a line or block) of LLM-generated source code and request that a variable name contained in the selection be altered. The same variable name may then be altered throughout the LLM-generated source code, both in the user's selection and elsewhere. In some such implementations, instances of the variable-to-be-altered that are found outside of the user's selection may be presented to the user one at a time, as a list, etc., so that the user can toggle through and approve (or reject) each proposed replacement."
Paragraph 31 "In some implementations, all or aspects of the NL based response system 120 can be implemented locally at the client device 110. In additional or alternative implementations, all or aspects of the NL based response system 120 can be implemented remotely from the client device 110 as depicted in FIG. 1 (e.g., at remote server(s)). In those implementations, the client device 110 and the NL based response system 120 can be communicatively coupled with each other via one or more networks 199, such as one or more wired or wireless local area networks (“LANs,” including Wi-Fi LANs, mesh networks, Bluetooth, near-field communication, etc.) or wide area networks (“WANs”, including the Internet)."
Paragraph 50 "Consistency engine 132 may be configured to evaluate the remainder of the raw LLM response outside of the subportion(s) extracted by selection extraction engine 130 in order maintain consistency between various aspects of the selected and unselected portions of the rendered LLM output. Suppose a user selects a middle paragraph of a scheduling email and issues the NL based request, “Please change the street number from 359 to 874.” The middle paragraph of the scheduling email (e.g., minus metadata instructions) may be extracted by selection extraction engine 130 and incorporated into a subsequent LLM input prompt by LLM input engine 126. This subsequent LLM input prompt may also include the user's request to change the street numbers. When the subsequent input request is processed by LLM response generation engine 128 using an LLM 129, the resulting LLM response may include the previous LLM response, except with the middle paragraph altered to reflect the new street numbers. However, if the street number were also included in another portion of the original rendered LLM output that the user didn't select, that other instance of the street number may not be replaced as requested, resulting in a scheduling email with inconsistent street numbers. Accordingly, in various implementations, consistency engine 132 may be configured to ensure that details changed within the selected portion of the original rendered LLM output (generated using the underlying raw LLM response) are also changed elsewhere, where applicable. In some implementations, consistency engine 132 may perform its actions heuristically, e.g., by extracting entities and facts from both the user selection and the remainder of the rendered LLM output and comparing them. In other implementations, the LLM 129 itself may be trained to maintain consistent factual details across both selected and unselected portions of LLM responses."
Paragraph 70 "At block 412, the system, e.g., by way of selection extraction engine 130, may extract or select a subportion (e.g., 258 in FIG. 2) of the first LLM response that corresponds to the selected subportion (e.g., 256A) of the first rendered LLM output (e.g., 254A), e.g., based on the indication received at block 410. At block 414, the system, e.g., by way of LLM input engine 126, may assemble, as a second LLM prompt, the selected subportion (e.g., 258) of the first LLM response (e.g., 252A) with data indicative of the request (e.g., 250B) to modify the selected subportion (258) of the first rendered LLM output. In some implementations, LLM input engine 126 may also include one or more implied requests, such as for consistency engine 132 to evaluate the various portions to ensure consistency between details and/or facts, and/or for the prior LLM response (e.g., 252A) outside of the selected portion (258) to be passed through, so that it can be included in the downstream LLM response (e.g., 252B in FIG. 2).")
Claim 7
Regarding Claim 7, Jacint in view of Agastya , further Jacint teach the method of claim 6, wherein at least a portion of the plurality of additional database entries were generated by the generative natural language model prior to being trained using the plurality of additional database entries.
(paragraph 25 " The indication of the subportion of the rendered LLM output that was selected by the user (e.g., starting and ending character positions) may be used to extract a portion of the original LLM response. This extracted portion may then be assembled into a follow up LLM prompt along with the user's follow up request. In some implementations, additional implied request(s) or command(s) may also be incorporated into the follow up LLM prompt that are designed to trigger selected aspects of the present disclosure. For instance, an additional implied request may be a request to “only modify the provided excerpt of the previous LLM response in accordance with the user's command. Leave the remainder of the previous LLM response unaltered.” In some implementations, the LLM may be trained and/or fine-tuned using implied requests such as these, along with LLM responses with subportions selected and corresponding user commands. This follow up LLM prompt may then be processed using the same LLM or a different LLM to generate a subsequent LLM response. The subportion of the subsequent LLM response that corresponds to the portion of the previously rendered LLM output that was selected by the user may be altered in accordance with the user's follow up request. In some implementations, the remainder of the subsequent LLM response may be left untouched, and thus may not necessarily be processed using the LLM, which conservers considerable resources."
Paragraph 56 "Starting at top right, a first request 250A may be received at user input engine 111, which in turn provides data indicative of the first request 250A (e.g., the request itself, embedding(s) generated therefrom, etc.) to LLM input engine 126 of NL based response system 120. First request 250A may be typed, may be transcribed using ASR on a spoken utterance, or may even be an implied query. Whichever the case, data indicative of first request 250A may be assembled by LLM input engine 126 into an LLM prompt (not depicted) that is then processed by LLM response generation engine 128 using one or more LLMs from database 129 to generate a first raw LLM response 252A. As noted previously, first raw LLM response 252A may include a sequence of tokens, such as a sequence of raw text that includes both content responsive to the request and metadata instructions interspersed therein. First raw LLM response 252A may be provided by UX engine 136 to rendering engine of client device 110. Rendering engine 112 may provide, e.g., a display and/or speakers, first rendered LLM output 254A, which may include various modalities of output, such as audible, images, text, etc."
Paragraph 48 "The LLM response generation engine 128 may be configured to apply one or more LLMs stored in an LLM database 129 to LLM input prompts generated by LLM input engine 126 to generate an LLM response. An LLM response may take various forms, such as a sequence of tokens that correspond to, represent, or directly convey words, phrases, embeddings, etc. LLMs stored in LLM database 129 may take a variety of form, such as PaLM, BARD, BERT, LaMDA, Meena, GPT, and/or any other LLM, such as any other LLM that is encoder-only based, decoder-only based, sequence-to-sequence based and that optionally includes an attention mechanism or other memory. Visual language models (VLMs) capable of processing images and text may be included as well."
Claim 9
Regarding Claim 9, Jacint in view of Agastya , further Agastya teaches
9. The method of claim 8, wherein the method further comprises, prior to obtaining the selection of the subset of the plurality of elements:
responsive to determining that the first element includes the first data input, determining a third textual prompt based on the first element;
(Paragraph 53 "Turning back to the method 200, consider, as yet another non-exhaustive example, the dashboard 800 shown in FIG. 8. In this example, the prompt 804 is “Where is DWA in state of CT” (where DWA is “designated wind area”). Here the user may then select the “submit” icon 810, as in S211, described above. The text component 127 determines the PII status for the prompt is PII-free in S212. The method then proceeds to S216 and a “No PII elements” output 902 (FIG. 9) is returned to the display in the Vetted Response field.")
generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents the input sanitization criterion for the first data input; and
(Paragraph 53 "Turning back to the method 200, consider, as yet another non-exhaustive example, the dashboard 800 shown in FIG. 8. In this example, the prompt 804 is “Where is DWA in state of CT” (where DWA is “designated wind area”). Here the user may then select the “submit” icon 810, as in S211, described above. The text component 127 determines the PII status for the prompt is PII-free in S212. The method then proceeds to S216 and a “No PII elements” output 902 (FIG. 9) is returned to the display in the Vetted Response field."
paragraph 56 "In a case the selected models are approved, the prompt may be transmitted to the AI via selection of the “Send to AI” icon 316 in S222. The prompt may be transmitted via a suitable Application Programming Interface (API). Prior to transmission of the prompt, the AI gateway tool 122: may append any additional prompts, as described above, to the prompt for transmission to the LLM, and may append any formatting parameters for the response to the prompt. The LLM output/response 1902 may be received, via a suitable API, at the AI response user interface display 1900 in S224, as shown in FIG. 19. The AI response display 1900 may also include the prompt 1904 as sent to the LLM, and the selected LLM (model) 1906. Pursuant to some embodiments, based on the response 1902, additional prompts may be sent to the LLM per the AI gateway tool 122. Further, data stores may be updated with the response 1902, and the response may then be used to update existing internal models and create new internal models.")
updating the database entry according to the third output to include the input sanitization criterion for the first data input.
(Paragraph 53 "Turning back to the method 200, consider, as yet another non-exhaustive example, the dashboard 800 shown in FIG. 8. In this example, the prompt 804 is “Where is DWA in state of CT” (where DWA is “designated wind area”). Here the user may then select the “submit” icon 810, as in S211, described above. The text component 127 determines the PII status for the prompt is PII-free in S212. The method then proceeds to S216 and a “No PII elements” output 902 (FIG. 9) is returned to the display in the Vetted Response field."
paragraph 56 "In a case the selected models are approved, the prompt may be transmitted to the AI via selection of the “Send to AI” icon 316 in S222. The prompt may be transmitted via a suitable Application Programming Interface (API). Prior to transmission of the prompt, the AI gateway tool 122: may append any additional prompts, as described above, to the prompt for transmission to the LLM, and may append any formatting parameters for the response to the prompt. The LLM output/response 1902 may be received, via a suitable API, at the AI response user interface display 1900 in S224, as shown in FIG. 19. The AI response display 1900 may also include the prompt 1904 as sent to the LLM, and the selected LLM (model) 1906. Pursuant to some embodiments, based on the response 1902, additional prompts may be sent to the LLM per the AI gateway tool 122. Further, data stores may be updated with the response 1902, and the response may then be used to update existing internal models and create new internal models.")
See claim 1 for rationale.
Claim 12
Regarding Claim 12 , Jacint in view of Agastya , further Jacint teach
12. The method of claim 1, wherein the first output includes a listing of the elements of the plurality of elements, and wherein the method further comprises, for each element of the plurality of elements of the database entry:
(Fig 3D (element 354B) shows the plurality of elements )
determining a third textual prompt based on the element, wherein the third textual prompt includes at least one of a request to re-generate the element, a request to generate an input sanitization criterion for the element, or a request to elaborate the element;
(Fig 3D (element 354B) shows the plurality of elements as well as selecting an element (element 356C) to regenerate in paragraph form (element 372) )
generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents an update to the element; and
(Paragraph 64 "In FIG. 3D, the user has selected a subportion 356C of the subsequent rendered LLM output 354B that identifies a list of activities that are planned for the birthday party. The user has provided, in query input field 372, the request, “Rewrite this in paragraph form.” Consequently, in FIG. 3E, another subsequent rendered LLM output 354C that includes a modified subportion 356D has been used to replace the selected subportion 356C. As requested, the modified subportion 356D describes, in paragraph form, the activities planned for the party.")
updating the element of the database entry according to the third output.
(Fig 3E shows the updated output (element 354C more specifically (element 356D)))
Claim 13
Regarding Claim 13 , Jacint in view of Agastya , further Jacint teach
13. The method of claim 1, wherein obtaining the selection of the subset of the plurality of elements includes receiving a command to re-generate the selected subset of the plurality of elements,
(Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372)
Fig 3C shows the new output where only the selected portion was changed (element 356B)
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
and wherein determining the second textual prompt based on the selected subset of the plurality of elements comprises determining the second textual prompt to include a portion of the first output that represents the selected subset of the plurality of elements
(Fig 3B shows the user selecting an element (in 354A (356A))
Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372)
Fig 3C shows the new output where only the selected portion was changed (element 356B)
)
and a request to re- generate the elements represented by the portion of the first output.
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
Claims 10 are rejected under 35 U.S.C. 102 (a)(2) as being anticipated by US Patent US 20250181824 A1, (Szabo; Jacint.), in view of US Patent US 20250131126 A1, (Kommanamanchi; Agastya.) in further view of US Patent US 11861321 B1, (O'Kelly; Brian.)
Claim 10
Regarding Claim 10, Jacint in view of Agastya, further Agastya teach receiving, from a user via a user interface, a textual description of the input sanitization criterion;
(Paragraph 46-49 " Initially, at S210, a prompt 304 is received. The prompt 304 may be received from a user 105 in the “Message to be Vetted” user entry field 302. In FIG. 3, the prompt is “Does phone number 5555555555 belong to Joe.” In other embodiments, the prompt may be received automatically and directly from an application 103, without user entry of text in a field.
The user then selects the “submit” 310 icon in S211.
In S212, a personal identifiable information (PII) status (e.g., “Contains PII” or “PII-free”) is determined. PII uses data to confirm an individual's identity. Sensitive PII may include, but is not limited to, a full name, face, home address, social security number, passport number, birthdate, driver's license, financial information, medical records, finger prints or handwriting sample, email address, phone number, etc. The AI gateway tool 122 identifies PII and other confidential text in the prompt 304 via the image component 125 and the text component 127. As described above, the image component 125 and the text component 127 may analyze the prompt via an internal ML model to identify PII or confidential information, or may access an external service to identify PII so the image component 127 and text component 127 may determine the PII status (e.g., presence or absence of PII data). Continuing with the non-exhaustive example of FIG. 3, the text component 127 may determine the PII status of the prompt 304.
In a case it is determined at S212, PII status is “Contains PII”, the method proceeds to S214 and a “Contains PII” output 402 (FIG. 4) is returned to the display in the Vetted Response field 306. The output 402 includes the PII status, and in the case of a “Contains PII” status, the output 402 further describes the PII included in the prompt. Additionally, in the case of the “Contains PII” status, the “Add model” icon 314 and “Send to AI” icon 316 are greyed out (as shown in FIG. 4) and not selectable by the user in response to the PII status of “Contains PII”."
paragraph 38 "The back-end application computer server 102 may also exchange information with a remote user device 124 (e.g., via a firewall 126). The back-end application computer server 102 may also exchange information via communication links 128 (e.g., via communication port 130 that may include a firewall) to communicate with different systems. The back-end application computer server 102 may also transmit information directly to an email server, workflow application, and/or calendar application 132 to facilitate automated communications and/or other actions. The back-end application computer server 102 may also transmit (via a firewall) information (e.g., prompts) to LLMs 134 after being approved by the AI gateway tool 122. According to some embodiments, an interactive graphical user interface platform of the back-end application computer server 102 may facilitate resource management, schedule recommendations, alerts, and/or the display of results via one or more remote administrator computers (e.g., to display the response to the prompt) and/or the remote user device 124. For example, the remote user device 124 may transmit a prompt and/or updated information regarding a record to the back-end application computer server 102. Based on the prompt/updated information, the back-end application computer server 102 may adjust data in the data store 104, and the change may (or may not) be used in connection with other systems. Note that the back-end application computer server 102 and/or any of the other devices and methods described herein may be associated with a third party, such as a vendor that performs a service for an enterprise (e.g., image processing, text processing).")
Jacint in view of Agastya do not explicitly teach all of determining a third textual prompt based on the textual description, wherein the third textual prompt includes a request for a regular expression representing the input sanitization criterion as described by the textual description; generating, via the natural language model, a third output based on the third textual prompt, wherein the third output includes the regular expression that represents the input sanitization criterion; and updating the database entry according to the third output to include the regular expression as the input sanitization criterion for the first data input.
However, Brian teaches
determining a third textual prompt based on the textual description, wherein the third textual prompt includes a request for a regular expression representing the input sanitization criterion as described by the textual description;
(Col 42 lines 56-67 and col 43 lines 0- 43 "A regular expression prompt template is determined at 1806. In some implementations, a regular expression prompt template may include at least two components. First, the regular expression prompt template may include one or more fillable portions that may be filled with text from a document to create a regular expression prompt. A fillable portion may be specified via a markup language. For instance, a fillable portion may include language such as <text portion>, which may be replaced with an actual text portion to create a regular expression prompt.
Second, the regular expression prompt template may include one or more natural language instructions instructing a large language model to generate one or more regular expressions. In some embodiments, the natural language instructions may be implemented in natural language, not computer code. The natural language instructions may include information such as a format to be used for generating the one or more regular expressions, an example of a regular expression to generate, and the like. The natural language instructions may also include other information, such as an instruction to associate a regular expression with a document structure level, a markup tag, or other such information.
An example of a regular expression prompt template that may be used to generate regular expressions is as follows. In the following example, the fillable portion “{% for clause in clauses %}<CC{{loop.index0}}>{{clause.text}}</CC{{loop.index0}}>{% endfor %}” indicates where to insert the input text portions to create the regular expression prompt from the regular expression prompt template. #Purpose You are an advanced legal AI assistant, proficient in understanding and generating an XML schema that outlines a contract. Your task is to analyze clauses from a contract and create an XML representation of the contract's structure, identifying how the sections and subsections are denoted and organized. ##Instructions Please consider the following instructions: 1. Examine the contract structure. Contracts generally use prefixing to indicate the hierarchy of sections. Use the provided clauses to determine: The number of layers in the contract The prefixes used to denote hierarchy levels Regex patterns that can be used to identify these prefixes 2. Format your response as XML tags. Each tag must have the following attributes: level: indicates a level of sectioning in the contract pattern: regex pattern that can be used to identify the prefix or formatting example: an example of the prefix" )
generating, via the natural language model, a third output based on the third textual prompt, wherein the third output includes the regular expression that represents the input sanitization criterion; and
(Col 44 lines 54-63 "
In some embodiments, multiple regular expression prompt templates may be generated. For instance, some or all of the text portions may be divided into different regular expression prompt templates, which may then be used independently to identify regular expressions.
The one or more regular expression prompts are transmitted to a large language model for completion at 1810. In some embodiments, the regular expression prompt may be transmitted to the large language model via the model API interface 252 shown in FIG. 2.")
updating the database entry according to the third output to include the regular expression as the input sanitization criterion for the first data input.
(Col 44 lines 54-63 "
In some embodiments, multiple regular expression prompt templates may be generated. For instance, some or all of the text portions may be divided into different regular expression prompt templates, which may then be used independently to identify regular expressions.
The one or more regular expression prompts are transmitted to a large language model for completion at 1810. In some embodiments, the regular expression prompt may be transmitted to the large language model via the model API interface 252 shown in FIG. 2.")
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Jacint in view of Agastya to incorporate the teachings of Brian to provide a “determining a third textual prompt based on the textual description, wherein the third textual prompt includes a request for a regular expression representing the input sanitization criterion as described by the textual description; generating, via the natural language model, a third output based on the third textual prompt, wherein the third output includes the regular expression that represents the input sanitization criterion; and updating the database entry according to the third output to include the regular expression as the input sanitization criterion for the first data input.” Doing so would make the system more accurate in the structure of text, as recognized by Brian. (col 47 lines 52-63)
Claims 11 are rejected under 35 U.S.C. 102 (a)(2) as being anticipated by US Patent US 20250181824 A1, (Szabo; Jacint.), in view of US Patent US 20250131126 A1, (Kommanamanchi; Agastya.) in further view of US Patent US 11861321 B1, (O'Kelly; Brian.) in further view of US Patent US 20240330365 A1, (Zawadowskiy; Andrew.)
Claim 11
Regarding Claim 11, Jacint in view of Agastya in further view of Brian do not explicitly teach all of receiving, from the user via the user interface, a test input; determining whether the test input satisfies the regular expression; and providing, via the user interface, an indication of whether the test input satisfies the regular expression.
However, Andrew. teach
11. The method of claim 10, further comprising:
receiving, from the user via the user interface, a test input;
(Paragraph 92 "In decision block 412, an inquire is performed to determine whether the coverage of regular expressions is sufficient? When the coverage of the regular expressions 108 is insufficient, the regular expressions 108 are updated using a new set of training log files 104. For example, the format of the log files can change over time, making it beneficial to learn the new format and learn new patterns for the regular expressions 108. Then the updated regular expressions 108 can be used in step 410 to parse the log files 110. The quality of the regular expressions 108 is reflected in how well the entities and relationships expressed in the log file 110 are recognized and extracted when performing the parsing."
Paragraph 93 "For example, the regular expression «\b[A-Z0-9._%-]+@[A-Z0-9.-]+\.[A-Z]{2.4}\b» should capture 100% of email address, and is therefore very good at recognizing and extracting email addresses. The regular expression «\b[A-Z0-9._%-]+@[A-Z0-9.-]+\.com» is less good for email addresses because it would only capture email addresses that end in .com“. Similarly, the effectiveness and sufficiency of the regular expressions 108 can be monitored in 412 using various metrics, such as what percentage of the characters in the log file 110 are matched by the regular expressions 108. Updating the regular expressions 108 can be triggered when one or more of these metrics falls below a predefined threshold.")
determining whether the test input satisfies the regular expression; and
(Paragraph 93 "For example, the regular expression «\b[A-Z0-9._%-]+@[A-Z0-9.-]+\.[A-Z]{2.4}\b» should capture 100% of email address, and is therefore very good at recognizing and extracting email addresses. The regular expression «\b[A-Z0-9._%-]+@[A-Z0-9.-]+\.com» is less good for email addresses because it would only capture email addresses that end in .com“. Similarly, the effectiveness and sufficiency of the regular expressions 108 can be monitored in 412 using various metrics, such as what percentage of the characters in the log file 110 are matched by the regular expressions 108. Updating the regular expressions 108 can be triggered when one or more of these metrics falls below a predefined threshold.")
providing, via the user interface, an indication of whether the test input satisfies the regular expression.
(Paragraph 98 "The decision block 412 can monitor the sufficiency of the regular expressions 108 with respect to both false positives and false negatives. Further, the decision block 412 can monitor the sufficiency of the regular expressions with respect to different categories and concepts in the log files. For example, if the statistical prevalence of timestamp entities is known for a particular format of log files, then a drastic decrease in the extraction of timestamp entities can be an indicator that the format of the timestamp entities has changed and the regular expression for the timestamp entities should be updated. Further, baseline statistics can be learned for different types/categories of entities and different types/categories of relationships. Then the statistics for these can be monitored and compared to the baseline to detect changes, which changes can indicate that format has changes to thereby trigger an update of the regular expressions 108.")
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Jacint in view of Agastya in further view of Brian to incorporate the teachings of Andrew to provide a “receiving, from the user via the user interface, a test input; determining whether the test input satisfies the regular expression; and providing, via the user interface, an indication of whether the test input satisfies the regular expression.” Doing so to see the sufficiency of the regular expression, as recognized by Andrew. (paragraph 98)
Claims 14, 15, 17, 18, 20 -23 are rejected under 35 U.S.C. 102 (a)(2) as being anticipated by US Patent US 20250181824 A1, (Szabo; Jacint.), in view of US Patent US 20250124066 A1, (McElvain; Gayle.).
Claim 14 and 18
Regarding Claim 14 and 18 Jacint teach
14. A non-transitory computer-readable medium, having stored thereon program instructions that, upon execution by a computing system, cause the computing system to perform operations comprising:
receiving, from a generative natural language model and based on a first textual prompt, a first output that represents a database entry, wherein the database entry includes a plurality of elements and wherein the first output includes a listing of the elements of the plurality of elements;
(Fig 3A shows the user entering the prompt (element 372) and the plurality of elements below (element 354A)
FIG 3c see element 356B, Fig 3 D element 356C, Elements 354A, 356B, and 356C would be the plurality of elements. Also the plurality of elements can be a single element based on the specification
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
for each element of the plurality of elements of the database entry:
(Fig 3A shows the user entering the prompt (element 372) and the plurality of elements below (element 354A)
FIG 3c see element 356B, Fig 3 D element 356C, Elements 354A, 356B, and 356C would be the plurality of elements. Also the plurality of elements can be a single element based on the specification)
updating the element of the database entry according to the third output;
(Fig 3A shows the user entering the prompt (element 372) and the plurality of elements below (element 354A)
FIG 3c see element 356B, Fig 3 D element 356C, Elements 354A, 356B, and 356C would be the plurality of elements. Also, the plurality of elements can be a single element based on the specification
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
obtaining a selection of a subset of the plurality of elements;
(Fig 3B shows the user selecting an element (in 354A (356A)))
determining a second textual prompt based on the selected subset of the plurality of elements; and
(Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372))
generating, via the generative natural language model, a second output based on the second textual prompt, wherein the second output represents an update to the database entry.
(Fig 3A shows the user entering the prompt (element 372) and the plurality of elements below (element 354A)
FIG 3c see element 356B, Fig 3 D element 356C, Elements 354A, 356B, and 356C would be the plurality of elements. Also, the plurality of elements can be a single element based on the specification
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
Jacint do not explicitly teach without user feedback, determining a third textual prompt based on the element,
wherein the third textual prompt includes at least one of a request to re-generate the element, a request to generate an input sanitization criterion for the element, or a request to elaborate the element;
generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents an update to the element; and
However, McElvain teach
without user feedback, determining a third textual prompt based on the element,
(Paragraph 28, provisional paragraph 21 "The LLM engine 120 may also receive or generate a prompt, which may be presented as input to an LLM of the LLM engine 120. The prompt may include information associated with the initial set of search results, the set of search criteria, or both. The prompt may include the input and portions of the initial set of search results identified as being relevant to the set of search criteria. The LLM engine 120 may output a response to the input based on content generated using the LLM of the LLM engine 120. The response may be generated using the LLM based on the prompt. In an aspect, the response may be generated via an iterative process. For example, during each iteration of the iterative process, a portion of the initial set of search results may be presented to the LLM and an interim response may be generated. The interim response and a next portion of the initial set of search results may then be provided as input to a next iteration of the iterative process until the response is output."
Since the model is going through an iterative process this would indicate that its without user feedback.)
wherein the third textual prompt includes at least one of a request to re-generate the element, a request to generate an input sanitization criterion for the element, or a request to elaborate the element; (the bold indicates what is being mapped to)
(Paragraph 28, provisional paragraph 21 "The LLM engine 120 may also receive or generate a prompt, which may be presented as input to an LLM of the LLM engine 120. The prompt may include information associated with the initial set of search results, the set of search criteria, or both. The prompt may include the input and portions of the initial set of search results identified as being relevant to the set of search criteria. The LLM engine 120 may output a response to the input based on content generated using the LLM of the LLM engine 120. The response may be generated using the LLM based on the prompt. In an aspect, the response may be generated via an iterative process. For example, during each iteration of the iterative process, a portion of the initial set of search results may be presented to the LLM and an interim response may be generated. The interim response and a next portion of the initial set of search results may then be provided as input to a next iteration of the iterative process until the response is output."
Since the response is being generated form the iteration this would indicate a re-generation.)
generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents an update to the element; and
(Paragraph 28, provisional paragraph 21 "The LLM engine 120 may also receive or generate a prompt, which may be presented as input to an LLM of the LLM engine 120. The prompt may include information associated with the initial set of search results, the set of search criteria, or both. The prompt may include the input and portions of the initial set of search results identified as being relevant to the set of search criteria. The LLM engine 120 may output a response to the input based on content generated using the LLM of the LLM engine 120. The response may be generated using the LLM based on the prompt. In an aspect, the response may be generated via an iterative process. For example, during each iteration of the iterative process, a portion of the initial set of search results may be presented to the LLM and an interim response may be generated. The interim response and a next portion of the initial set of search results may then be provided as input to a next iteration of the iterative process until the response is output."
Since the response is being regenerated this would indicate that the output is being updated.)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Jacint to incorporate the teachings of McElvain to provide a “without user feedback, determining a third textual prompt based on the element, wherein the third textual prompt includes at least one of a request to re-generate the element, a request to generate an input sanitization criterion for the element, or a request to elaborate the element; generating, via the generative natural language model, a third output based on the third textual prompt, wherein the third output represents an update to the element; and” Doing so would Alter the format of the response, evaluate the accuracy of the response, and incorporate negative treatment information to the response, as recognized by McElvain. (Paragraph 8).
Further regarding claim 18, Jacint in view of McElvain, further Jacint teach
one or more processors; and
(Paragraph 42 " Further, the client device 110, the NL based response system 120, and/or the search system 140 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199. In some implementations, one or more of the software applications can be installed locally at the client device 110, whereas in other implementations one or more of the software applications can be hosted remotely (e.g., by one or more servers) and can be accessible by the client device 110 over one or more of the networks 199.")
memory, containing program instructions that, upon execution by the one or more processors, cause the system to perform operations comprising:
(Paragraph 42 " Further, the client device 110, the NL based response system 120, and/or the search system 140 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199. In some implementations, one or more of the software applications can be installed locally at the client device 110, whereas in other implementations one or more of the software applications can be hosted remotely (e.g., by one or more servers) and can be accessible by the client device 110 over one or more of the networks 199.")
Claim 15
Regarding Claim 15, Jacint in view of McElvain, further Jacint teach the non-transitory computer-readable medium of claim 14, wherein the generative natural language model has been trained using a plurality of additional database entries related to operation of a particular managed network,
(Paragraph 27 "Accordingly, in some implementations, the LLM may be trained and/or fine-tuned to process commands to account for discrepancies between facts or details contained inside and outside of a user's selection. For example, an explicit user follow up request to replace a first date with a second date in a selected portion of rendered LLM output (generated from an underlying LLM response) may trigger generation of an implied request to also replace the first date with the second date elsewhere in the rendered LLM output, even in portion(s) not selected by the user. This feature may be particularly useful when the LLM is used to generate complex structured language such as source code, mathematical proofs, etc. For instance, a user may select a particular code segment (e.g., a line or block) of LLM-generated source code and request that a variable name contained in the selection be altered. The same variable name may then be altered throughout the LLM-generated source code, both in the user's selection and elsewhere. In some such implementations, instances of the variable-to-be-altered that are found outside of the user's selection may be presented to the user one at a time, as a list, etc., so that the user can toggle through and approve (or reject) each proposed replacement."
Paragraph 31 "In some implementations, all or aspects of the NL based response system 120 can be implemented locally at the client device 110. In additional or alternative implementations, all or aspects of the NL based response system 120 can be implemented remotely from the client device 110 as depicted in FIG. 1 (e.g., at remote server(s)). In those implementations, the client device 110 and the NL based response system 120 can be communicatively coupled with each other via one or more networks 199, such as one or more wired or wireless local area networks (“LANs,” including Wi-Fi LANs, mesh networks, Bluetooth, near-field communication, etc.) or wide area networks (“WANs”, including the Internet)."
Paragraph 50 "Consistency engine 132 may be configured to evaluate the remainder of the raw LLM response outside of the subportion(s) extracted by selection extraction engine 130 in order maintain consistency between various aspects of the selected and unselected portions of the rendered LLM output. Suppose a user selects a middle paragraph of a scheduling email and issues the NL based request, “Please change the street number from 359 to 874.” The middle paragraph of the scheduling email (e.g., minus metadata instructions) may be extracted by selection extraction engine 130 and incorporated into a subsequent LLM input prompt by LLM input engine 126. This subsequent LLM input prompt may also include the user's request to change the street numbers. When the subsequent input request is processed by LLM response generation engine 128 using an LLM 129, the resulting LLM response may include the previous LLM response, except with the middle paragraph altered to reflect the new street numbers. However, if the street number were also included in another portion of the original rendered LLM output that the user didn't select, that other instance of the street number may not be replaced as requested, resulting in a scheduling email with inconsistent street numbers. Accordingly, in various implementations, consistency engine 132 may be configured to ensure that details changed within the selected portion of the original rendered LLM output (generated using the underlying raw LLM response) are also changed elsewhere, where applicable. In some implementations, consistency engine 132 may perform its actions heuristically, e.g., by extracting entities and facts from both the user selection and the remainder of the rendered LLM output and comparing them. In other implementations, the LLM 129 itself may be trained to maintain consistent factual details across both selected and unselected portions of LLM responses."
Paragraph 70 "At block 412, the system, e.g., by way of selection extraction engine 130, may extract or select a subportion (e.g., 258 in FIG. 2) of the first LLM response that corresponds to the selected subportion (e.g., 256A) of the first rendered LLM output (e.g., 254A), e.g., based on the indication received at block 410. At block 414, the system, e.g., by way of LLM input engine 126, may assemble, as a second LLM prompt, the selected subportion (e.g., 258) of the first LLM response (e.g., 252A) with data indicative of the request (e.g., 250B) to modify the selected subportion (258) of the first rendered LLM output. In some implementations, LLM input engine 126 may also include one or more implied requests, such as for consistency engine 132 to evaluate the various portions to ensure consistency between details and/or facts, and/or for the prior LLM response (e.g., 252A) outside of the selected portion (258) to be passed through, so that it can be included in the downstream LLM response (e.g., 252B in FIG. 2).")
and wherein at least a portion of the plurality of additional database entries were generated by the natural language model prior to being trained using the plurality of additional database entries.
(paragraph 25 " The indication of the subportion of the rendered LLM output that was selected by the user (e.g., starting and ending character positions) may be used to extract a portion of the original LLM response. This extracted portion may then be assembled into a follow up LLM prompt along with the user's follow up request. In some implementations, additional implied request(s) or command(s) may also be incorporated into the follow up LLM prompt that are designed to trigger selected aspects of the present disclosure. For instance, an additional implied request may be a request to “only modify the provided excerpt of the previous LLM response in accordance with the user's command. Leave the remainder of the previous LLM response unaltered.” In some implementations, the LLM may be trained and/or fine-tuned using implied requests such as these, along with LLM responses with subportions selected and corresponding user commands. This follow up LLM prompt may then be processed using the same LLM or a different LLM to generate a subsequent LLM response. The subportion of the subsequent LLM response that corresponds to the portion of the previously rendered LLM output that was selected by the user may be altered in accordance with the user's follow up request. In some implementations, the remainder of the subsequent LLM response may be left untouched, and thus may not necessarily be processed using the LLM, which conservers considerable resources."
Paragraph 56 "Starting at top right, a first request 250A may be received at user input engine 111, which in turn provides data indicative of the first request 250A (e.g., the request itself, embedding(s) generated therefrom, etc.) to LLM input engine 126 of NL based response system 120. First request 250A may be typed, may be transcribed using ASR on a spoken utterance, or may even be an implied query. Whichever the case, data indicative of first request 250A may be assembled by LLM input engine 126 into an LLM prompt (not depicted) that is then processed by LLM response generation engine 128 using one or more LLMs from database 129 to generate a first raw LLM response 252A. As noted previously, first raw LLM response 252A may include a sequence of tokens, such as a sequence of raw text that includes both content responsive to the request and metadata instructions interspersed therein. First raw LLM response 252A may be provided by UX engine 136 to rendering engine of client device 110. Rendering engine 112 may provide, e.g., a display and/or speakers, first rendered LLM output 254A, which may include various modalities of output, such as audible, images, text, etc."
Paragraph 48 "The LLM response generation engine 128 may be configured to apply one or more LLMs stored in an LLM database 129 to LLM input prompts generated by LLM input engine 126 to generate an LLM response. An LLM response may take various forms, such as a sequence of tokens that correspond to, represent, or directly convey words, phrases, embeddings, etc. LLMs stored in LLM database 129 may take a variety of form, such as PaLM, BARD, BERT, LaMDA, Meena, GPT, and/or any other LLM, such as any other LLM that is encoder-only based, decoder-only based, sequence-to-sequence based and that optionally includes an attention mechanism or other memory. Visual language models (VLMs) capable of processing images and text may be included as well." )
Claim 17 and 20
Regarding Claim 17 and 20 , Jacint in view of McElvain , further Jacint teach
13. The method of claim 1, wherein obtaining the selection of the subset of the plurality of elements includes receiving a command to re-generate the selected subset of the plurality of elements,
(Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372)
Fig 3C shows the new output where only the selected portion was changed (element 356B)
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
and wherein determining the second textual prompt based on the selected subset of the plurality of elements comprises determining the second textual prompt to include a portion of the first output that represents the selected subset of the plurality of elements
(Fig 3B shows the user selecting an element (in 354A (356A))
Fig 3B shows the user entering a second prompt where he wants it to start at a different time (element 372)
Fig 3C shows the new output where only the selected portion was changed (element 356B)
)
and a request to re- generate the elements represented by the portion of the first output.
Paragraph 18 "Implementations are described herein for using LLMs to modify selected subportions—i.e., less than the entirety—of LLM outputs. More particularly, but not exclusively, implementations are described herein for determining which subportion(s) of LLM outputs have been selected by a user, and modifying those selected subportion(s) based on a request from the user to generate modified versions of those selected subportion(s) of the LLM output. A user may select a subportion of an LLM output in various ways, such as highlighting content (text and/or images) using a pointer device, touchscreen, and/or keyboard, verbally identifying a particular portion (e.g., “shorten the second paragraph,” “update the map to give driving directions instead of subway directions”), and so forth.")
Claim 21
Regarding Claim 21 , Jacint in view of McElvain , further Jacint teach The system of claim 18, wherein the generative natural language model has been trained using a plurality of additional database entries related to operation of a particular managed network, and wherein at least a portion of the plurality of additional database entries were
(Paragraph 27 "Accordingly, in some implementations, the LLM may be trained and/or fine-tuned to process commands to account for discrepancies between facts or details contained inside and outside of a user's selection. For example, an explicit user follow up request to replace a first date with a second date in a selected portion of rendered LLM output (generated from an underlying LLM response) may trigger generation of an implied request to also replace the first date with the second date elsewhere in the rendered LLM output, even in portion(s) not selected by the user. This feature may be particularly useful when the LLM is used to generate complex structured language such as source code, mathematical proofs, etc. For instance, a user may select a particular code segment (e.g., a line or block) of LLM-generated source code and request that a variable name contained in the selection be altered. The same variable name may then be altered throughout the LLM-generated source code, both in the user's selection and elsewhere. In some such implementations, instances of the variable-to-be-altered that are found outside of the user's selection may be presented to the user one at a time, as a list, etc., so that the user can toggle through and approve (or reject) each proposed replacement."
Paragraph 31 "In some implementations, all or aspects of the NL based response system 120 can be implemented locally at the client device 110. In additional or alternative implementations, all or aspects of the NL based response system 120 can be implemented remotely from the client device 110 as depicted in FIG. 1 (e.g., at remote server(s)). In those implementations, the client device 110 and the NL based response system 120 can be communicatively coupled with each other via one or more networks 199, such as one or more wired or wireless local area networks (“LANs,” including Wi-Fi LANs, mesh networks, Bluetooth, near-field communication, etc.) or wide area networks (“WANs”, including the Internet)."
Paragraph 50 "Consistency engine 132 may be configured to evaluate the remainder of the raw LLM response outside of the subportion(s) extracted by selection extraction engine 130 in order maintain consistency between various aspects of the selected and unselected portions of the rendered LLM output. Suppose a user selects a middle paragraph of a scheduling email and issues the NL based request, “Please change the street number from 359 to 874.” The middle paragraph of the scheduling email (e.g., minus metadata instructions) may be extracted by selection extraction engine 130 and incorporated into a subsequent LLM input prompt by LLM input engine 126. This subsequent LLM input prompt may also include the user's request to change the street numbers. When the subsequent input request is processed by LLM response generation engine 128 using an LLM 129, the resulting LLM response may include the previous LLM response, except with the middle paragraph altered to reflect the new street numbers. However, if the street number were also included in another portion of the original rendered LLM output that the user didn't select, that other instance of the street number may not be replaced as requested, resulting in a scheduling email with inconsistent street numbers. Accordingly, in various implementations, consistency engine 132 may be configured to ensure that details changed within the selected portion of the original rendered LLM output (generated using the underlying raw LLM response) are also changed elsewhere, where applicable. In some implementations, consistency engine 132 may perform its actions heuristically, e.g., by extracting entities and facts from both the user selection and the remainder of the rendered LLM output and comparing them. In other implementations, the LLM 129 itself may be trained to maintain consistent factual details across both selected and unselected portions of LLM responses."
Paragraph 70 "At block 412, the system, e.g., by way of selection extraction engine 130, may extract or select a subportion (e.g., 258 in FIG. 2) of the first LLM response that corresponds to the selected subportion (e.g., 256A) of the first rendered LLM output (e.g., 254A), e.g., based on the indication received at block 410. At block 414, the system, e.g., by way of LLM input engine 126, may assemble, as a second LLM prompt, the selected subportion (e.g., 258) of the first LLM response (e.g., 252A) with data indicative of the request (e.g., 250B) to modify the selected subportion (258) of the first rendered LLM output. In some implementations, LLM input engine 126 may also include one or more implied requests, such as for consistency engine 132 to evaluate the various portions to ensure consistency between details and/or facts, and/or for the prior LLM response (e.g., 252A) outside of the selected portion (258) to be passed through, so that it can be included in the downstream LLM response (e.g., 252B in FIG. 2).")
generated by the generative natural language model prior to being trained using the plurality of additional database entries.
(paragraph 25 " The indication of the subportion of the rendered LLM output that was selected by the user (e.g., starting and ending character positions) may be used to extract a portion of the original LLM response. This extracted portion may then be assembled into a follow up LLM prompt along with the user's follow up request. In some implementations, additional implied request(s) or command(s) may also be incorporated into the follow up LLM prompt that are designed to trigger selected aspects of the present disclosure. For instance, an additional implied request may be a request to “only modify the provided excerpt of the previous LLM response in accordance with the user's command. Leave the remainder of the previous LLM response unaltered.” In some implementations, the LLM may be trained and/or fine-tuned using implied requests such as these, along with LLM responses with subportions selected and corresponding user commands. This follow up LLM prompt may then be processed using the same LLM or a different LLM to generate a subsequent LLM response. The subportion of the subsequent LLM response that corresponds to the portion of the previously rendered LLM output that was selected by the user may be altered in accordance with the user's follow up request. In some implementations, the remainder of the subsequent LLM response may be left untouched, and thus may not necessarily be processed using the LLM, which conservers considerable resources."
Paragraph 56 "Starting at top right, a first request 250A may be received at user input engine 111, which in turn provides data indicative of the first request 250A (e.g., the request itself, embedding(s) generated therefrom, etc.) to LLM input engine 126 of NL based response system 120. First request 250A may be typed, may be transcribed using ASR on a spoken utterance, or may even be an implied query. Whichever the case, data indicative of first request 250A may be assembled by LLM input engine 126 into an LLM prompt (not depicted) that is then processed by LLM response generation engine 128 using one or more LLMs from database 129 to generate a first raw LLM response 252A. As noted previously, first raw LLM response 252A may include a sequence of tokens, such as a sequence of raw text that includes both content responsive to the request and metadata instructions interspersed therein. First raw LLM response 252A may be provided by UX engine 136 to rendering engine of client device 110. Rendering engine 112 may provide, e.g., a display and/or speakers, first rendered LLM output 254A, which may include various modalities of output, such as audible, images, text, etc."
Paragraph 48 "The LLM response generation engine 128 may be configured to apply one or more LLMs stored in an LLM database 129 to LLM input prompts generated by LLM input engine 126 to generate an LLM response. An LLM response may take various forms, such as a sequence of tokens that correspond to, represent, or directly convey words, phrases, embeddings, etc. LLMs stored in LLM database 129 may take a variety of form, such as PaLM, BARD, BERT, LaMDA, Meena, GPT, and/or any other LLM, such as any other LLM that is encoder-only based, decoder-only based, sequence-to-sequence based and that optionally includes an attention mechanism or other memory. Visual language models (VLMs) capable of processing images and text may be included as well." )
Claim 22
Regarding Claim 22, Jacint in view of McElvain , further Jacint teach 22. (New) The system of claim 18, wherein the second output represents an update to the selected subset of the plurality of elements of the database entry.
(Fig 3c shows the new output where only the selected portion was changed (element 356B))
Claim 23
Regarding Claim 23 , Jacint in view of McElvain , further Jacint teach 23. (New) The non-transitory computer-readable medium of claim 14, wherein the second output represents an update to the selected subset of the plurality of elements of the database entry.
(Fig 3c shows the new output where only the selected portion was changed (element 356B))
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ALI M HASSAN whose telephone number is (571)272-5331. The examiner can normally be reached Monday - Friday 8:00am - 4:00pm.
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, Paras Shah can be reached at (571)270-1650. 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.
/ALI M HASSAN/Examiner, Art Unit 2653
/Paras D Shah/Supervisory Patent Examiner, Art Unit 2653
04/18/2026
Prosecution Insights