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
Applicants’ amendment filed on 1/29/26 has been entered. Claims 1, 5, 10 and 14 have been amended. No claims have been canceled. No new claims have been added. Claims 1-20 are still pending in this application, with claims 1, 5 and 14 being independent.
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 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 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-10, 12-20 are rejected under 35 U.S.C. 103 as being unpatentable over Non-Patent Literature “Interactive and Visual Prompt Engineering for Ad-hoc Task Adaptation with Large Language Models” to Strobelt et al. (“Strobelt”) in view of U.S. Patent Application Publication No. 20210209513 to Torres et al. (“Torres”).
As to claims 1, 5 and 14, Strobelt discloses a system, a method and one or more non-transitory, computer-readable storage media, the method comprising:
receiving, by a prompt development system [Abstract: “PromptlDE allows users to experiment with prompt variations, visualize prompt performance, and iteratively optimize prompts”], a request to perform further adaptation that tunes performance of a pre-trained natural language processing (NLP) machine learning (ML) model [abstract: "newly created ad-hoc models"; section 3, first paragraph: "Prompting is typically used in conjunction with large pre-trained language model"; as seen in section 5.3 users may add template variables to retrieved and copied examples and generate different prompt variation] that performs an NLP task according to an input prompt [section 1: "prompting [ ... ] has become popular for developing ad-hoc end-user tasks in NLP"; section 3, first paragraph: "solving ad-hoc NLP tasks" using "prompting"] selected from a prompt task collection maintained by the prompt development system [refer to figures 1 and 5 which shows that different prompt variations from a collection can be seen; figure 8 and section 5.2 show that templates can be added to a cart and exported and deployed in the PromptlDE repository; moreover the PromptSource shopping cart of figure 6 shows templates that have been created by the community and may serve as inspiration] ; generating, by the prompt development system, an adaption job to perform the requested further adaptation of the pre-trained NLP ML model using an adaption data set specified by the request [figure 6 and section 5.3: the user may generate adapted prompt variations; these are then carried out, a corresponding data set has been defined using the data browser]; evaluating, by the prompt development system, performance of the adaptation job that changes performance of the pre-trained NLP ML model using the plurality of examples based on providing the input prompt to the adapted pre-trained NLP ML model to generate a result of the adaptation job [figure 6 shows that the refined prompts are evaluated against ground truths; this is also seen in section 5.3]; and providing, by the prompt development system, the result of the adaptation job [Fig. 6].
Strobelt does not expressly disclose the adaptation data set comprising a plurality of examples of the NLP task.
In the same or similar field of invention, Torres discloses the adaptation data set comprising a plurality of examples of the NLP task [Torres paragraphs 0019 : “multi-task tuned model 252 may ingest multi-task input data 212 including requests for multiple types of NLP predictions as input…”, paragraph 0038, also see Fig. 5]. Torres also discloses optimizing performance of the adaptation job that changes performance of the pre-trained NLP ML model using the plurality of examples [Torres paragraphs 0019, 0038, 0039-0053].
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify Strobelt to have features of the adaptation data set comprising a plurality of examples of the NLP task as taught by Torres. The suggestion/motivation would have been to improve the efficiency of machine learning systems serving NLP predictions at scale by decreasing the amount of system down time and reducing the number of hours the system operates below capacity by using the Adapter Service architecture [Torres paragraph 0025].
As to claims 2, 6 and 15, Strobelt discloses wherein the adaptation job performs a prompt tuning technique to tune the pre-trained NLP ML model [Fig. 6, page 5, also see rejection of claims 1, 5 and 14].
As to claims 3, 7 and 16, Strobelt discloses wherein adaptation job performs an in-context learning technique to tune the pre-trained NLP ML model [Fig. 6, page 5, also see rejection of claims 1, 5 and 14].
As to claims 4, 12 and 20, Strobelt discloses wherein the prompt development system is a machine learning service offered by a provider network, and wherein the prompt development system is further configured to: receive, via the interface, a request to deploy the tuned NLP ML model; provision a host for the tuned NLP ML model; and provide, via the interface, a model endpoint for a client application to submit inference requests to the host to generate respective inferences using the tuned NLP ML model [Sections 4.1 and 6.2].
As to claims 8 and 17, Strobelt discloses wherein the adaptation job performs a fine-tuning technique to tune the pre-trained NLP ML model [Fig. 6, page 5, also see rejection of claims 5 and 14].
As to claims 9 and 18, Strobelt discloses wherein the pre-trained NLP ML model was included in a prompt recommendation provided in response to a discovery request performed by the prompt development system [Figs 3, 6, page 5, also see rejection of claims 5 and 14, examples including trained models can be searched and retrieved by the user].
As to claim 10, Strobelt discloses wherein evaluating performance of the adaptation job that tunes performance of the pre-trained NLP ML model comprises: generating using the adapted NLP ML model one or more inferences for input test data in accordance with the input prompt; and determining inference performance for the one or more inferences based on ground truth labels for the input test data [Fig. 6].
Claims 11 is rejected under 35 U.S.C. 103 as being unpatentable over Non-Patent Literature “Interactive and Visual Prompt Engineering for Ad-hoc Task Adaptation with Large Language Models” to Strobelt et al. (“Strobelt”) and U.S. Patent Application Publication No. 20210209513 to Torres et al. (“Torres”) in further view of U.S. Patent Application Publication No. 20190279038 to Nicol (“Nicol”).
As to claim 11, Strobelt and Torres disclose the method of claim 5 [see rejection of claim 5].
Further, Strobelt discloses wherein the result of the adaptation job comprises computational performance [Fig. 6 page 2-3, 5].
Strobelt and Torres do not expressly disclose wherein the result of the adaptation job comprises computational performance and inference performance.
In the same or similar field of invention, Nicol discloses wherein the result of the adaptation job comprises computational performance and inference performance [Nicol paragraphs 0009, 0047, 0050].
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify Strobelt and Torres to have features of wherein the result of the adaptation job comprises computational performance and inference performance as taught by Nicol. The suggestion/motivation would have been to provide a data flow graph, which is useful visual representation for understanding a variety of highly complex computing tasks [Nicol paragraph 0009].
Response to Arguments
Applicant’s arguments with respect to claims 1-20 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.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANTIM G SHAH whose telephone number is (571)270-5214. The examiner can normally be reached Mon-Fri 7:30am-4pm.
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, Ahmad Matar can be reached at 571-272-7488. 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.
/ANTIM G SHAH/Primary Examiner, Art Unit 2693