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
Claims 1, 2, 5, 6, 9, 10, and 13-20 have been amended. Claims 6 and 8 are canceled and claims 21-22 are newly presented. Claims 1-5, 7, and 9-22 are presented for examination.
Response to Arguments
Applicant’s arguments filed on 6/26/2026 have been reviewed. Following are the responses to the amendments.
Rejection under 35 U.S.C. 101
Applicant arguments have been considered and are persuasive, hence the rejection under
35 U.S.C. 101 is withdrawn.
Rejection under 35 U.S.C. 103
(II). Rejection of Claims 1-2, 4-5, 10-11, and 15 under 35 U.S.C. § 103
Applicant states that “the combination of Sharma and Najib fails to disclose at least the amended subject matter of claim 1”. Examiner agrees that the previously applied combination does not disclose the amended limitations. Accordingly, a new ground of rejection has been entered based on Sharma in view of Sanderson and Chen.
(III). Rejection of Claim 3 under 35 U.S.C. § 103
Applicant argues that claim 3 depends from claim 1 and that Kumare fails to cure the alleged deficiencies of the prior combination as applied to claim 1. The examiner agrees with Applicant’s argument. Accordingly, a new ground of rejection has been entered for claim 3 based on the newly applied combination discussed above together with Khumbare.
(IV). Rejection of Claims 6 and 13-14 under 35 U.S.C. @ 103
Applicant correctly notes that the prior office action inconsistently referred to Hajarnis and Khumbare. The examiner acknowledges the inconsistency. Claim 6 has been canceled, and with respect to claims 13-14 a new ground of rejection has been entered based on the newly applied combination discussed above.
(V). Rejection of Claim 7 under 35 U.S.C. @ 103
Applicant argues that claim 7 should be allowable because it depends from claim 1 and the additional reference does not cure the alleged deficiencies of the prior rejection of claim 1. The examiner agrees with Applicant’s argument regarding the prior rejection. Accordingly, a new ground of rejection has been entered for claim 7 based on the newly applied combination discussed above and the additional teachings relied upon for claim 7.
(VI). Rejection of Claim 8 under 35 U.S.C. @ 103
Claim 8 has been canceled. Accordingly, the rejection is moot.
(VII). Rejection of Claim 9 under 35 U.S.C. @ 103
Applicant argues that claim 9 should be allowable because it depends from claim 1 and the additional reference does not cure the alleged deficiencies of the prior rejection of claim 1. The examiner agrees with Applicant’s argument regarding the prior rejection. Accordingly, a new ground of rejection has been entered for claim 9 based on the newly applied combination discussed above and the additional teachings relied upon for claim 9.
(VIII). Rejection of Claims 12, 16, 17, 18, 19, 20 under 35 U.S.C. @ 103
Applicant argues that claim 12 and 16-20 should be allowable because they depend from claims 10 and 15 and the additional references do not cure the alleged deficiencies of the prior rejection of claims 10 and 15. The examiner agrees with Applicant’s argument regarding the prior rejection. Accordingly, a new ground of rejection has been entered for claims 12 and 16-20 based on the newly applied combination discussed above and the additional teachings relied upon for claim 12 and 16-20.
Claim Objections
Claim 22 is objected to because of the following informalities: "the moderator approval data" lacks proper antecedent basis. Claim 15 does not previously recite moderator approval data.. Appropriate correction is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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, 2, 4, 5, 10, 13, 14 15 are rejected under 35 U.S.C. 103 as being unpatentable over Sharma et al. (US 20250322168 A1) in view of Najib et al. (WO 2025000074 A1), Sanderson et al. (US20250335699) and Chen et al. (US 20250252301 A1).
Sharma teaches A system, comprising: at least one processor; and at least one memory that stores executable instructions that, when executed by the at least one processor, facilitate performance of operations, comprising (Fig. 6, demonstrates the overall system architecture): performing the inferences with the large language model (), comprising: analyzing first text that is received based on first user input data (Para 0004, “delivering the input text to an inclusive prompt recommendation model as the text is being received, the inclusive prompt recommendation model being trained to process the input text to: determine at least one
of an intent and a context”), and context of the first text, the analyzing using a large language model to identify a first recommendation to alter the first text (Fig 4, Teaches the use of the
model to recommend additional words to add to text. This comprises an alteration.) to satisfy an inclusive-language criterion (Para 0004,“generate at least one inclusive prompt recommendation
based at least in part on the determined at least one of the intent and context” and where the
criterion is taught in the abstract, “The system can include an ethical filtering mechanism for
ensuring that prompt recommendations do not have language that directly or indirectly promotes
bias and/or stereotypes.”), performing a third tuning of the large language model based on the user feedback data, to produce an updated large language model (Para 0023, “The training system uses training data based on user interactions and feedback pertaining the use of the system which has been collected over time. This training refines the model over time and can improve the system's understanding and performance.”, this tuning occurs after the model is deployed, meaning the tunings as taught by Sanderson and Sharma would have already occurred making this analogous to a third tuning); analyzing second text received based on second user input data with the updated large language model to identify a second recommendation to alter the second text to satisfy the inclusive-language criterion. (Sharma teaches that the system can be
improved over time [Para 0023, “This training refines the model over time and can improve the
system's understanding and performance.”] which implies the ability to receive a
recommendation multiple times and repeat the process which is already taught for the first
recommendation)
Sharma does not teach receiving user feedback data based on the first recommendation.
However, Najib teaches receiving user feedback data based on the first recommendation
(Para 0175, “the method 400 includes receiving review submissions from customers, the review
submissions initiated after each customer transaction”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Najib to gain the benefit of more
opportunities to encourage customers to share their feedback (Para 0175, “triggering review
prompts that encourage customers to share their feedback.”).
Sharma modified by Najib does not teach before performing inferences with a large language model, performing a first tuning of the large language model with a group of pairs, wherein respective pairs of the group of pairs comprise respective exclusive language examples and corresponding expected inclusive language examples, performing a third tuning of the large language model based on the user feedback data, to produce an updated large language model; analyzing second text received based on second user input data with the updated large language model to identify a second recommendation to alter the second text to satisfy the inclusive-language criterion.
However, Sanderson teaches before performing inferences with a large language model, performing a first tuning of the large language model with a group of pairs (Para 0099, “the set of data instances include multiple pairs of (prompt, positive text)”, and ) and para 0100, “the training module 330 trains or further fine-tunes parameters of the machine learned models” wherein the prompt/positive-text pairs correspond to the claimed group of pairs and the fine tuning corresponds to the first tuning) wherein respective pairs of the group of pairs comprise respective exclusive language examples and corresponding expected inclusive language examples (Para 0064, “rewrite the sentence to replace any detected phrases”, and para 0066, “remove references to the employee’s race; they are not relevant” wherein the original text containing the detected biased or non-inclusive language corresponds to the claimed exclusive language example, and the rewritten text corresponds to the inclusive example).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Sharma in order to incorporate the teachings of Sanderson in order to improve the model’s ability to generate appropriate examples (Para 0064).
Sharma modified by Sanderson does not teach after the first tuning of the large language model with the group of pairs, performing a second tuning of the large language model with a low-rank adaptation of a defined large language models process.
However, Chen teaches after the first tuning of the large language model with the group of pairs, performing a second tuning of the large language model with a low-rank adaptation of a defined large language models process (Para 0026, “For a pre-trained weight matrix … LoRA models the weight update” and para 0032, “during a second phase of fine-tuning, the direction component is indirectly updated through LoRA” where Chen’s LoRA fine tuning corresponds to the claimed second tuning, and Chen teaches applying LoRA during a second phase of fine-tuning after an earlier tuning phase).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Sharma in order to incorporate the teachings of Sanderson in order to adapt the already trained model to reduce the number of parameters that must be updated during fine-tuning (Para 0026).
Regarding claim 2, Sharma modified by Sanderson and Chen teaches a third tuning. (Para 0023, “The training system uses training data based on user interactions and feedback pertaining the use of the system which has been collected over time. This training refines the model over time and can improve the system's understanding and performance.”, this tuning occurs after the model is deployed, meaning the tunings as taught by Sanderson and Sharma would have already occurred making this analogous to a third tuning)
Sharma does not teach the system wherein the operations further
comprise: receiving moderator approval data that is indicative of the user feedback data being
approved by a monitor before performing the tuning of the language model based on the user
feedback data.
However, Najib teaches the system wherein the operations further comprise: receiving
moderator approval data that is indicative of the user feedback data being approved by a monitor
before performing the tuning of the language model based on the user feedback data. (Para 0163,
“the outcomes of the moderation (i.e., reviews being approved or declined) are fed back into the
device 300 to further refine the learning models.”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Najib to gain the benefit of the
system being able to evolve in order to incorporate new data (Para 0086, “ensures that the
algorithms performed or deployed at or by the Al-based analysis module 122 continue to evolve
in response to new data”).
Regarding claim 4, Sharma does not teach the system wherein the moderator approval
data is first moderator approval data, wherein the user feedback data is first user feedback data
and wherein the operations further comprise: refraining from updating the large language model
based on receiving second moderator approval data that is indicative of the second user feedback
data being rejected.
However, Najib teaches the system wherein the moderator approval data is first
moderator approval data, wherein the user feedback data is first user feedback data. (Para 0163,
The system of moderation is taught in “the outcomes of the moderation (i.e., reviews
being approved or declined) are fed back into the device 300 to further refine the
learning models.”, during the first iteration of the process it would include the first moderation
and first user feedback); and wherein the operations further comprise: refraining from updating
the large language model based on receiving second moderator approval data that is indicative of
the second user feedback data being rejected (Para 0148, “The decision-making module 340
automates the approval of reviews that align with moderation policies and rejects those that do
not”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Najib to gain the benefit of the
system being able to evolve in order to incorporate new data (Para 0086, “ensures that the
algorithms performed or deployed at or by the Al-based analysis module 122 continue to evolve
in response to new data”).
Regarding claim 5, Sharma does not teach the system wherein the tuning of the large
language model is performed based on the receiving of the moderator approval data.
However, Najib teaches the system wherein the tuning of the large language model is
performed based on the receiving of the moderator approval data (Para 0086, “where the
outcomes of the moderation (i.e., reviews being approved or declined) are fed back into the
system 100 to further refine the learning models”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Najib to gain the benefit of the
system being able to evolve in order to incorporate new data (Para 0086, “ensures that the
algorithms performed or deployed at or by the Al-based analysis module 122 continue to evolve
in response to new data”).
Regarding claim 11, Sharma teaches the method comprising: iteratively updating, by the
system, the large language model based on a group of user feedback data that comprises the user
feedback data (Para 0023, “The training system uses training data based on user interactions and
feedback pertaining the use of the system which has been collected over time.
This training refines the model over time and can improve the system's understanding and
performance.”).
Regarding claim 13, Sharma does not teach updating, by the system, the pairs offline.
However, Sanderson teaches updating, by the system, the pairs offline. (Para 0101, “the newly obtained feedback may be used to reconstruct the training dataset” wherein reconstructing the stored training dataset using newly obtained feedback corresponds to updating the previously created prompt/positive text pairs offline))
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Sharma in order to incorporate the teachings of Sanderson in order to improve the training dataset as additional feedback is obtained (Para 0101).
Regarding claim 14, Sharma does not teach wherein updating the pairs offline produces updated pairs, and wherein the tuning of the large language model comprises: inputting the updated pairs into the large language model.
However, Sanderson teaches wherein updating the pairs offline produces updated pairs (Para 0101, “the newly obtained feedback may be used to reconstruct the training dataset”, wherein reconstructing the training dataset with newly obtained feedback produces an updated set of prompt/positive-text training pairs corresponding to the claimed updated pairs), and wherein the tuning of the large language model comprises: inputting the updated pairs into the large language model (Para 0100, “the training module 330 trains or further fine tunes parameters of the machine learned models … based on the created training dataset” wherein using the reconstructed training dataset to further fine tune the model corresponds to inputting the updated pairs into the large language model during the claimed third tuning).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Sharma in order to incorporate the teachings of Sanderson in order to refine the model using the updated training data (Para 0100).
Claims 10 and 15 are analogous to claim 1 in that they recite substantially the same limitations. They are therefore rejected for the same reasons as stated above.
Claim 21 is analogous to claim 2 in that it recites substantially the same limitations. It is therefore rejected for the same reasons.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Sharma et al. (US 20250322168 A1) in view of Najib (WO 2025000074 A1) Sanderson (US20250335699) and Chen (US 20250252301 A1) as applied to claims 1, 2, 4, 5, 10, 15 above, and further in view of Mariko et al. (US 20250298958 A1).
Sharma does not teach the system wherein the large language model is tuned to specialize in text classification or text-to-text generation.
However, Mariko teaches the system wherein the large language model is tuned to
specialize in text classification or text-to-text generation. (Para 0027, “Fine-tuned or domain-
specific models are LLMs that have undergone additional training on domain-specific data to
improve their performance in particular areas or with particular tasks like text classification and
language generation”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Mariko to gain the benefit of
improving their performance in particular areas (Para 0027).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Sharma et al. (US 20250322168 A1) in view of Najib (WO 2025000074 A1) Sanderson (US20250335699) and Chen (US 20250252301 A1) as applied to claims 1, 2, 4, 5, 10, 15 above, and further in view of Kaan et al. (WO 2025199345 A1).
Sharma modified by Najib does not teach the system wherein the operations further
comprise: before analyzing the first text that is received based on the first user input data with
the large language model, performing a fourth tuning of the large language model via providing a multi-shot prompt as input to the large language model, wherein the multi-shot prompt comprises a description of an intent to suggest inclusive language and an output that is to be output by the large language model.
However, Kaan teaches performing a fourth tuning of the large language model via providing a multi-shot prompt as input to the large language model, wherein the multi-shot prompt comprises a description of an intent to suggest inclusive language and an output that is to be output by the large language model. (Pg. 9 Ln 8: “For instance, a system prompt with fine-tuning examples could be as follows: System prompt: ‘You will receive a stream of text, your task is to determine if someone is talking to you, or if it’s ambient conversation, and then extract the user’s intent. Do not answer their question and always respond in valid JSON format.’ Example input: ‘The weather is nice today, but how will it be tomorrow?’ Example output:”, this teaches a description of intent (determine if someone is talking to you), and an output (Example output) When incorporated into Sharma as modified by Sanderson and Chen, which already performes the claimed first, second, and third tunings, Kaan’s additional prompt based tuning corresponds to the claimed fourth tuning).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in order to incorporate the teachings of Kaan to gain the benefit of a
more natural and intuitive user experience (Abstract, “This approach facilitates a more natural
and intuitive user experience”).
Claim 12, 16, 17, 18, 19, 20 is rejected under 35 U.S.C. 103 as being unpatentable over
Sharma et al. (US 20250322168 A1) in view of Najib (WO 2025000074 A1) Sanderson (US20250335699) and Chen (US 20250252301 A1) as applied to claims 1, 2, 4, 5, 10, 15 above, and further in view of Seth et al. (US 20250342630 A1).
Regarding claim 12, Sharma modified by Najib does not teach providing, by the system,
the first recommendation via a plugin to an email program.
However, Seth does teach providing, by the system, the first recommendation via a
plugin to an email program. (Para 0052, “For example, the system can work on the web or within a virtual meeting and collaboration application (e.g., Microsoft Teams®) or an email application
(e.g., Outlook®)” where the output (the first recommendation) is rendered via an email
application).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output. (Para 0052).
Regarding claim 16, Sharma modified by Najib does not teach sending, by the system,
the first recommendation to be rendered via a word processor program.
However, Seth does teach sending, by the system, the first recommendation to be
rendered via a word processor program. (Para 0052, “Such applications can be a stand-alone
applications, a plug-in or an Edit button of any application on the client device 105, such as the
browser application 112, the native application 114, and the like” where an enterprise
management program falls under being an application on the client device).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output (Para 0052).
Regarding claim 17, Sharma modified by Najib does not teach sending, by the system,
the first recommendation to be rendered via a team collaboration application.
However, Seth does teach sending, by the system, the first recommendation to be
rendered via a team collaboration application. (Para 0052, “For example, the system can work on the web or within a virtual meeting and collaboration application (e.g., Microsoft Teams®).”
where Microsoft Teams comprises a team collaboration application).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output (Para 0052).
Regarding claim 18, Sharma modified by Najib does not teach sending, by the system,
the first recommendation to be rendered via an enterprise management program.
However, Seth does teach sending, by the system, the first recommendation to be
rendered via an enterprise management program. (Para 0052, “Such applications can be a stand-
alone applications, a plug-in or an Edit button of any application on the client device 105, such
as the browser application 112, the native application 114, and the like” where an enterprise
management program falls under being an application on the client device).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output. (Para 0052).
Regarding claim 19, Sharma modified by Najib does not teach sending, by the system,
the first recommendation to be rendered via an enterprise social networking service .
However, Seth does teach sending, by the system, the first recommendation to be
rendered via an enterprise social networking service (Para 0052, “The system can also work
within a social media website/application (e.g., Facebook®, Instagram®)”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output. (Para 0052).
Regarding claim 20, Sharma modified by Najib does not teach sending, by the system,
the first recommendation to be rendered via a wiki service.
However, Seth does teach sending, by the system, the first recommendation to be
rendered via a wiki service (Para 0052, “Such applications can be a stand-alone applications, a
plug-in or an Edit button of any application on the client device 105, such as the browser
application 112, the native application 114, and the like” where a wiki service is interpreted to
compose a browser application.”).
It would have been obvious to one of ordinary skill in the art to modify Sharma before
the effective filing date in such a way as to incorporate the teachings of Seth in order to allow
more options to render the output.
Allowable Subject Matter
Claims 3, 22 are objected to as being dependent upon a rejected base claim, but would be
allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
THIS ACTION IS MADE FINAL. 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 MICHAEL ALAN FOSTER JR. whose telephone number is (571)272-8874. The examiner can normally be reached M - F 8:00am - 5:00pm, Alternate Fridays Off.
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, Hai Phan can be reached at (571) 272-6338. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MICHAEL A FOSTER JR/Examiner, Art Unit 2654
/HAI PHAN/Supervisory Patent Examiner, Art Unit 2654
Prosecution Insights