DETAILED ACTION
Introduction
1. This office action is in response to Applicant’s submission filed on 2/24/2025. Claims 1-6 are pending in the application and have been examined.
Notice of Pre-AIA or AIA Status
2. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
3. The information disclosure statement (IDS) submitted on 3/10/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
4. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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.
5. Claims 1-6 are rejected under 35 U.S.C. 103 as unpatentable over U.S. Pat. App. Pub. No. 20230195428 (Deng et al., hereinafter “Deng”) in view of U.S. Pat. App. Pub. No. 20250111553 (Shen et al., hereinafter “Shen”).
With regard to Claim 1, Deng describes:
“A process for mitigating bias in a large language model (LLM), the process comprising:
evolving the LLM using a continuous learning (CL) process by sequentially exposing the LLM to evolving datasets DI, D2,...DT, each of which corresponds to a different domain, wherein at each timestep t, the LLM is updated using only a current dataset Dt, resulting in updated LLMs, LLMZ, LLM2,... LLMT after each sequential exposure; (Paragraph 90 describes that the model is input a different, evolving data set at each timestep.)
tuning each LLMT, LLM2,... LLMT after each sequential exposure on a downstream task Tskj, Tsk2,...TskT related to its specific domain Dt; (Paragraph 20 describes that the model is tuned based on the desired task.)
after each tuning of each LLM, LLM2,... LLMT, applying
i. multiple behavioral assessment tasks to each updated LLMT, LLM2,... LLMT to evaluate [[bias and]] natural language understanding and linguistics, and (Paragraph 30 describes that supervised metrics are calculated for the tuned models.)
ii. a domain-specific model performance task for evaluating performance of each LLM, LLM2,... LLMT on its corresponding downstream task Tsk, Tsk2,... Tskr; (Paragraph 30 describes that a set of code quality factors is used to evaluate the performance of the model in carrying out the task.)
determining correlations between the multiple behavioral assessment tasks; (Paragraph 13 describes that the supervised metrics are scored multiple ways to determine the task performance.)
Deng does not explicitly describe evaluating bias, or
“determining statistically significant correlations between different bias evaluation scores and one or more natural language understanding and linguistics scores, wherein the determining further includes identifying one or more natural language understanding and linguistics tasks underlying each statistically significant correlation and the respective bias;
identifying one or more biases in the LLM and ascertaining the identified one or more natural language understanding and linguistics tasks underlying the one or more biases as determined in the statistically significant correlations; and
augmenting the LLM to account for the identified one or more natural language understanding and linguistics tasks underlying the one or more biases and mitigate the one or more biases when the LLM performs downstream tasks Tsk, Tsk2,...TskT.”
However, Shen describes:
“determining statistically significant correlations between different bias evaluation scores and one or more natural language understanding and linguistics scores, wherein the determining further includes identifying one or more natural language understanding and linguistics tasks underlying each statistically significant correlation and the respective bias; (Paragraph 139 describes that bias evaluation scores are correlated to text encoder scores to finetune the model and the language understanding components.)
identifying one or more biases in the LLM and ascertaining the identified one or more natural language understanding and linguistics tasks underlying the one or more biases as determined in the statistically significant correlations; and (Paragraphs 128 and 129 describe determine bias levels for different biases.)
augmenting the LLM to account for the identified one or more natural language understanding and linguistics tasks underlying the one or more biases and mitigate the one or more biases when the LLM performs downstream tasks Tsk, Tsk2,...TskT.” (Paragraph 139 describes finetuning as few as five soft tokens as a prompt prefix to largely reduce gender bias as “augmenting the LLM.”)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the bias calculations as described by Shen into the system of Deng to provide bias mitigation, as described in paragraph 139 of Shen.
Deng does not explicitly describe the subject matter of Claim 2. However, Shen describes “determining correlations between the multiple behavioral assessment tasks includes creating multiple unique metric evaluations by combining the multiple behavioral assessment tasks and the domain-specific model performance tasks and analyzing the multiple unique metric evaluations to determine correlations therebetween when applied to the updated LLMZ, LLM2,... LLMT at checkpoints Cpti, Cpt2,... Cptr.” Paragraphs 128 and 129 describes that multiple tasks (gender, race) are analyzed to determine the various bias scores.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the bias tasks/scores as described by Shen into the system of Deng to provide bias mitigation, as described in paragraph 139 of Shen.
Deng does not explicitly describe the subject matter of Claim 3. However, Shen describes “the multiple behavioral assessment tasks evaluate two or more of gender bias, religious bias, and race bias.” Paragraph 128 describes evaluating gender and race bias.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the bias evaluations as described by Shen into the system of Deng to provide bias mitigation, as described in paragraph 139 of Shen.
Deng does not explicitly describe the subject matter of Claim 6. However, Shen describes “a domain-specific model performance score for each LLMI, LLM2,... LLMT on its corresponding downstream task Tskj, Tsk2,... TskT is statistically unchanged by augmenting the LLM to account for the identified one or more natural language understanding and linguistics tasks underlying the one or more biases and mitigate the one or more biases.” Paragraph 139 describes that the text encode is fine tuned while the U-Net is kept unchanged.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include the bias evaluations as described by Shen into the system of Deng to provide bias mitigation, as described in paragraph 139 of Shen.
Allowable Subject Matter
6. Claims 4 and 5 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.
The prior art of record does not teach or suggest “augmenting the LLM includes adding or subtracting one or more task vectors generated in accordance with the identified one or more natural language understanding and linguistics tasks underlying the one or more biases to mitigate the one or more identified biases of the LLM when performing one or more downstream tasks Tski, Tsk2,...TskT” as recited in Claim 4.
Conclusion
7. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
U.S. Pat. App. Pub. No. 20250190768 (Grayson et al.) describes a device that trains a model at each of a plurality of timesteps, and then tunes it for specific tasks.
8. Any inquiry concerning this communication or earlier communications from the examiner should be directed to EDWARD TRACY whose telephone number is (571)272-8332. The examiner can normally be reached Monday-Friday 9 AM- 5PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Bhavesh Mehta can be reached on 571-272-7453. 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.
/EDWARD TRACY JR./Examiner, Art Unit 2656
