DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This Office Action is sent in response to Applicant's Communication received on July 28, 2025 for application number 19/151,590. This Office hereby acknowledges receipt of the following and placed of record in file: Specification, Drawings, Abstract, Oath/Declaration, and Claims.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 01/13/2026 is noted. The submission 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 § 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 1-4, 6-7, 9-12, 14-15, and 17-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to Judicial Exceptions without significantly more. The claims recite mathematical relationships, mathematical formulas or equations, mathematical calculation and a mental process. This judicial exception is not integrated into a practical application because the recitation of generic computer and generic computer components does not sufficient to integrate the recited judicial exception into a practical application. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the claims only recites generic computer components, which are well-understood, routine, and conventional.
Revised Patent Subject Matter Eligibility Guidance
The USPTO has published revised guidance on the application of § 101. USPTO’s 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (Jan. 7, 2019) (“Guidance”). Under the Guidance, the Examiner first look to whether the claim recites:
(1) any judicial exceptions, including certain groupings of abstract ideas (i.e., mathematical concepts, certain methods of organizing human activity such as a fundamental economic practice, or mental processes) (Guidance, Step 2A, prong 1); and
(2) additional elements that integrate the judicial exception into a practical application (see Manual of Patent Examining Procedure (MPEP) § 2106.05(a)-(c), (e)-(h) (9th Ed., Rev. 08.2017, 2018)) (Guidance, Step 2A, prong 2).
Only if a claim (1) recites a judicial exception and (2) does not integrate that exception into a practical application, do the Examiner then look to whether the claim:
(3) adds a specific limitation beyond the judicial exception that is not “well-understood, routine, conventional” in the field (see MPEP § 2106.05(d)); or
(4) simply appends well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception. (Guidance (Step 2B)).
Evaluate Step 2A Prong One
(a) identify the specific limitation(s) in the claim that recites an abstract idea;
(b) determine whether the identified limitation(s) falls within at least one of the groupings of abstract ideas enumerated in the 2019 Revised Patent Subject Matter Eligibility Guidance.
In TABLE 1 below, the Examiner identifies in italics the specific claim limitations that recite an abstract idea.
TABLE 1
Independent Claim 1
Analysis Under Revised Guidance
(a) A computer-implemented method for recycling of task-specific prompts for machine-learned models, the method comprising: obtaining, by a computing system comprising one or more computing devices, a task- specific prompt for a machine-learned model, wherein the task-specific prompt is indicative of a task of a plurality of tasks the machine-learned model is configured to perform; determining, by the computing system, a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine-learned model; and based at least in part on the difference, modifying, by the computing system, the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model.
(b) determining, by the computing system, a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine-learned model
“determining a difference between a base version of the machine-learned model and an updated version of the machine-learned model” is an abstract idea, i.e., “a mathematical calculation” to calculate/measure the value or probability between the two different machine-learned models.
(c) modifying, by the computing system, the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model
“modifying the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model …” is an abstract idea, i.e., “a mathematical calculation” or “mathematical formula”, to adjust the task-specific prompt to correspond to an updated version of the machine-learned model.
In view of the above analysis, Claims 1, 9, and 17 recites an abstract idea under the Revised Guidance because the limitations (b) – (c) each recite mathematical relationship, mathematical calculation and/or a mental process. Dependent claims 2-4, 6-7, 10-12, 14-15, and 18-20 also recite abstract idea because they include limitations (b) – (c) by virtue of their dependencies to claims 1, 9, and 17.
Dependent claims 2-4, 6-7, 10-12, 14-15, and 18-20 further recites additional limitations. However, these limitations also recite abstract idea, i.e., “mathematical concept – mathematical formulas or equations, mathematical calculations” and i.e., a “mental process” similar to the limitations of claims 1, 9, and 17, discussed above.
Evaluate Step 2A Prong Two:
Evaluate whether the claim as a whole integrated the recited Judicial exception into a Practical Application of the exception.
Having determined that the claims recites a judicial exception, the analysis under the Guidance turns now to determining whether there are “additional element that integrate the judicial exception into a practical application”. The examiner determines whether the recited judicial exception is integrated into a practical application that exception by: (1) identifying whether there are any additional elements recited in the claim beyond the judicial exceptions; and (2) evaluating those additional elements individually and in combination to determine whether they integrate the exception into a practical application”.
Independent claims 1, 9, and 17 further recite “one or more computing devices”, “machine- learned models”, “one or more processors”, and “one or more non-transitory computer-readable media”, which is a generic/conventional computer storage. Claims 1, 9, and 17 do not recite any additional element that integrate the judicial exception into a practical application. The recitation of generic computer and generic computer components does not sufficient to integrate the recited judicial exception into a practical application. Guidance at MPEP 2106.04 (“Performance of a claim limitation using generic computer components does not necessarily preclude the claim limitation from being in the mathematical concepts grouping.”)
As discussed above, independent claims 1, 9, and 17 recites the mathematic calculation steps to determining a difference between a base version of the machine-learned model and an updated version of the machine-learned model, and based at least in part on the difference, modifying, by the computing system, the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model. These limitations are processes that, under broadest reasonable interpretation, covers performance of the limitation in a mathematical process, but for the recitations of generic computer components. That is, other than reciting a “one or more computing devices”, “machine- learned models”, “one or more processors”, and “one or more non-transitory computer-readable media”, nothing in the claim element precludes the step from practically being performed in a human mind or with the aid of pen and paper. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mathematical concept, then it falls within the “Mathematical Concepts” grouping of abstract ideas (mathematical relationships, mathematical formulas or equations, mathematical calculations (see MPEP § 2106.04(a)(2), subsection I)).
Evaluate Step 2B:
Evaluate whether the claim provide an inventive concept, i.e., does the claim recite additional element(s) or a combination of elements that amount to significantly more than the judicial exception in the claim?
At Step 2B, the evaluation of the insignificant extra-solution activity consideration takes into account whether or not the extra-solution activity is well-known. See MPEP 2106.05(g). The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim does not add any specific limitations beyond what is well-understood, routine, and conventional. Here, claims 1, 9, and 17 recite “one or more computing devices”, “machine- learned models”, “one or more processors”, and “one or more non-transitory computer-readable media”, which are mere generic computer components that are recited at a high level of generality, and, as disclosed in the specification, is also well-understood, routine, conventional activity when expressed at this high level of generality. Mere instructions to apply an exception using a generic computer component cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B. Therefore, the claims do not provide an inventive concept (significantly more than the abstract idea) and is not eligible.
These additional elements are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer components. Further, the claim recitations of obtaining data.
Obtaining is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity to the judicial exception with no evidence of improvement. Accordingly, the additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea, thus fail to integrate the abstract idea into a practical application. See MPEP 2106.05(g).
The claim does 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 elements of obtaining data (receiving or transmitting over a network), are well-understood, routine and conventional activity according to MPEP 2106.05(d)(II)(i), thus, cannot provide an inventive concept.
As a result, representative claim(s) 1, 9, and 17 does not recite any elements, or ordered combination of elements, which transforms the abstract idea into a patent-eligible subject matter. In addition, the claim(s) does not recite (i) an improvement to the functionality of a computer or other technology or technical field (see MPEP 2106.05(a); (ii) a “particular machine” to apply or use the judicial exception (see MPEP 2106.05(b); (iii) a particular transformation of an article to a different state or thing (see 2106.05(c). Further, the claim does not recite any improvement to computer functionality or specify how the one or more processors are used to improve functionality of a computing device. Considering the claim(s) as a whole, the additional elements fail to apply or use the abstract idea in a meaningful way and the additional limitations recited beyond the judicial exception itself fail to integrate the exception into a practical application. Accordingly, the claims 1-20 of this application are rejected.
Claims 2, 10, and 18 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) the abstract idea of a mental process, for example the claims are directed toward the Mathematical Concept of trained large language models, under broadest reasonable interpretation, covers performance of the limitation in calculations but for the recitation of generic computer components. The limitations associated with trained large language models are considered to be an abstract idea that falls in the “Mathematical Concept” grouping of abstract ideas.
Claims 3-4, 11-12, and 19-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) the abstract idea of a mathematical concept, for example the claims are directed toward the mathematical calculation for determining a difference between vocabulary embeddings of machine-learned models and linear combination including task specific prompts, under broadest reasonable interpretation, covers performance of the limitation in calculations but for the recitation of generic computer components. The limitations associated with determining differences between vocabulary embeddings of machine learned models and linear combination including task specific prompts are considered to be an abstract idea that falls in the “Mathematical Concept” grouping of abstract ideas.
Claims 6-7 and 14-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) the abstract idea of a mathematical concept, for example the claims are directed toward the mathematical calculation for evaluating loss function for a difference between task specific prompts and generating a learned transformation matrix, under broadest reasonable interpretation, covers performance of the limitation in calculations but for the recitation of generic computer components. The limitations associated with evaluating loss function for a difference between task specific prompts and generating a learned transformation matrix are considered to be an abstract idea that falls in the “Mathematical Concept” grouping of abstract ideas.
The claim(s) 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 a processor to perform the determining and modifying steps amounts to no more than mere instructions to apply the exception using a generic computer component. The limitations related to obtaining are considered by the examiner to be well-understood, routine and conventional activity according to MPEP 2106.05(d)(II)(i), because the inventive subject matter is directed toward machine learning model prompt tuining. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Because of these reasons the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim(s) 1-4, 6-7, 9-12, 14-15, and 17-20 are rejected.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 6-7 and 14-15 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for pre-AIA the inventor(s), at the time the application was filed, had possession of the claimed invention, had possession of the claimed invention.
Regarding claims 6 and 14 recite “… a ground-truth task-specific prompt…”. Examiner has read the entire discloser to ascertain clear support for limitations. Examiner did not find any support in written description support regarding a ground-truth task-specific prompt. It is unclear if this is a typographical error.
Regarding claims 7 and 15 recite “… a learned transformation matrix…”. Examiner has read the entire discloser to ascertain clear support for limitations. Examiner did not find any support in written description support regarding a learned transformation matrix. It is unclear if this is a typographical error.
Therefore, in view of the foregoing analyses, and Examiner’s reading of the entire disclosure of Applicant’s instant invention, Examiner finds a lack of written description support for the claim language: “… a ground-truth task-specific prompt…” and “…a learned transformation matrix …”. Applicant is invited to show where support for the limitations can be found within the disclosure of the invention.
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 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 9-10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Newman et al. (US 2023/0083512)(hereinafter Newman) in view of Reza et al. (US 2023/0237277)(hereinafter Reza).
Regarding claim 1, Newman teaches a computer-implemented method for recycling of task-specific prompts for machine-learned models, the method comprising: obtaining, by a computing system comprising one or more computing devices, a task- specific prompt for a machine-learned model (see Figs. 7-8, para [0063-0064], discloses obtaining natural language prompt (task-specific prompts) to perform a task in a language model (machine learning model)), wherein the task-specific prompt is indicative of a task of a plurality of tasks the machine-learned model is configured to perform (see para [0048], discloses the natural language prompt indicating a task for a BERT Base Cased model fine-tuned on a 41-way classification task).
Newman does not explicitly teach determining, by the computing system, a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine-learned model; and based at least in part on the difference, modifying, by the computing system, the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model.
Reza teaches determining, by the computing system, a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine-learned model (see Fig. 1, para [0009-0011], para [0033], discloses determining differences between text labels and text predicted for mask tokens in model parameters of a model (base version model) and updated model parameters for a machine learning language model (updated version of machine-learned model) for predicting a solution for a task); and based at least in part on the difference, modifying, by the computing system, the task- specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model (see Figs. 2-3, para [0061-0062], para [0077-0078], discloses based on differences in prompting functions, modifying prompt with dynamic prompting (updated task-specific prompt) that corresponds to respective text in a prompting template, creating updated model parameters in a machine learning language model (updated version of machine-learned model) in fine-tuning process).
Newman/Reza are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman to include an updated version of machine-learned model with an updated task-specific prompt from disclosure of Reza. The motivation to combine these arts is disclosed by Reza as “improving prompt-based learning by dynamically developing a contextual set of prompts based on relevant aspects extracted” (para [0028]) and include an updated version of machine-learned model with an updated task-specific prompt is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claim 9, Newman teaches a computing system for recycling of task-specific prompts for machine- learned models, comprising: one or more processors; and one or more non-transitory computer-readable media that store instructions that (see para [0058], discloses processor and medium), when executed by the one or more processors, cause the computing system to perform operations, the operations comprising: obtaining a task-specific prompt for a machine-learned model (see Figs. 7-8, para [0063-0064], discloses obtaining natural language prompt (task-specific prompts) to perform a task in a language model (machine learning model)), wherein the task- specific prompt is indicative of a task of a plurality of tasks the machine-learned model is configured to perform (see para [0048], discloses the natural language prompt indicating a task for a BERT Base Cased model fine-tuned on a 41-way classification task).
Newman does not explicitly teach determining a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine- learned model; and based at least in part on the difference, modifying the task-specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model.
Reza teaches determining a difference between a base version of the machine-learned model and an updated version of the machine-learned model different than the base version of the machine- learned model (see Fig. 1, para [0009-0011], para [0033], discloses determining differences between text labels and text predicted for mask tokens in model parameters of a model (base version model) and updated model parameters for a machine learning language model (updated version of machine-learned model) for predicting a solution for a task); and based at least in part on the difference, modifying the task-specific prompt to obtain an updated task-specific prompt that corresponds to the updated version of the machine-learned model (see Figs. 2-3, para [0061-0062], para [0077-0078], discloses based on differences in prompting functions, modifying prompt with dynamic prompting (updated task-specific prompt) that corresponds to respective text in a prompting template, creating updated model parameters in a machine learning language model (updated version of machine-learned model) in fine-tuning process).
Newman/Reza are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman to include an updated version of machine-learned model with an updated task-specific prompt from disclosure of Reza. The motivation to combine these arts is disclosed by Reza as “improving prompt-based learning by dynamically developing a contextual set of prompts based on relevant aspects extracted” (para [0028]) and include an updated version of machine-learned model with an updated task-specific prompt is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claim 17, Newman teaches one or more non-transitory computer-readable media that store instructions that, when executed by one or more processors, cause the one or more processors to perform operations (see para [0058], discloses processor and medium), the operations comprising: obtaining a task-specific prompt for a first machine-learned model (see Figs. 7-8, para [0063-0064], discloses obtaining natural language prompt (task-specific prompts) to perform a task in a language model (machine learning model)), wherein the task- specific prompt is indicative of a task of a plurality of tasks the first machine-learned model is configured to perform (see para [0048], discloses the natural language prompt indicating a task for a BERT Base Cased model fine-tuned on a 41-way classification task).
Newman does not explicitly teach determining a difference between the first machine-learned model and a second machine- learned model different than the first machine-learned model; and based at least in part on the difference, modifying the task-specific prompt to obtain an updated task-specific prompt that corresponds to the second machine-learned model.
Reza teaches determining a difference between the first machine-learned model and a second machine- learned model different than the first machine-learned model; (see Fig. 1, para [0009-0011], para [0033], discloses determining differences between text labels and text predicted for mask tokens in model parameters of a model (base version model) and updated model parameters for a machine learning language model (updated version of machine-learned model) for predicting a solution for a task); and based at least in part on the difference, modifying the task-specific prompt to obtain an updated task-specific prompt that corresponds to the second machine-learned model (see Figs. 2-3, para [0061-0062], para [0077-0078], discloses based on differences in prompting functions, modifying prompt with dynamic prompting (updated task-specific prompt) that corresponds to respective text in a prompting template, creating updated model parameters in a machine learning language model (updated version of machine-learned model) in fine-tuning process).
Newman/Reza are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman to include an updated version of machine-learned model with an updated task-specific prompt from disclosure of Reza. The motivation to combine these arts is disclosed by Reza as “improving prompt-based learning by dynamically developing a contextual set of prompts based on relevant aspects extracted” (para [0028]) and include an updated version of machine-learned model with an updated task-specific prompt is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claims 2 and 10, Newman/Reza teach a method of claim 1 and a system of claim 9.
Newman further teaches wherein the machine-learned model comprises a trained large language model (see para [0020], para [0022], discloses trained large language model such as BERT, GPT, and Gopher).
Claims 3-4, 11-12, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Newman et al. (US 2023/0083512)(hereinafter Newman) in view of Reza et al. (US 2023/0237277)(hereinafter Reza) as applied to claims 1, 9, and 17, and in further view of Drain et al. (US 2022/0066914)(hereinafter Drain).
Regarding claims 3 and 11, Newman/Reza teach a method of claim 1 and a system of claim 9.
Newman/Reza do not explicitly teach wherein determining the difference between the base version of the machine-learned model and the updated version of the machine-learned model comprises: determining, by the computing system, a difference between vocabulary embeddings of the base version of the machine-learned model and vocabulary embeddings of the updated version of the machine-learned model.
Drain teaches wherein determining the difference between the base version of the machine-learned model and the updated version of the machine-learned model comprises: determining, by the computing system, a difference between vocabulary embeddings of the base version of the machine-learned model and vocabulary embeddings of the updated version of the machine-learned model (see Figs. 2-3, Fig. 5, para [0007-0008], para [0034, 0041], discloses determining semantic and statistical differences between a pre-trained neural transformer model and a fine-tuned neural transformer model).
Newman/Reza/Drain are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman/Reza to determine a difference between vocabulary embeddings from disclosure of Drain. The motivation to combine these arts is disclosed by Drain as “in order to learn the semantics and statistical properties of the natural language” (para [0007]) and determining a difference between vocabulary embeddings is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claims 4 and 12, Newman/Reza teach a method of claim 1 and a system of claim 9.
Newman/Reza do not explicitly teach determining the difference between the base version of the machine-learned model and the updated version of the machine-learned model comprises determining, by the computing system, a first linear combination of vocabulary embeddings of the base version of the machine-learned model and the task-specific prompt; and wherein modifying the task-specific prompt comprises determining, by the computing system, the updated task-specific prompt based at least in part on the first linear combination and vocabulary embeddings of the updated version of the machine-learned model.
Drain teaches determining the difference between the base version of the machine-learned model and the updated version of the machine-learned model comprises determining, by the computing system, a first linear combination of vocabulary embeddings of the base version of the machine-learned model and the task-specific prompt (see Fig. 3, Fig. 10, para [0008], para [0086], discloses fine-tuning of neural transformer model, including assert statement prediction and task-specific head of model); and wherein modifying the task-specific prompt comprises determining, by the computing system, the updated task-specific prompt based at least in part on the first linear combination and vocabulary embeddings of the updated version of the machine-learned model (see Fig. 3, Fig. 10, para [0008], para [0041], discloses fine-tuning neural transformer model with attention with test-assert triplets).
Newman/Reza/Drain are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman/Reza to determine a difference between vocabulary embeddings from disclosure of Drain. The motivation to combine these arts is disclosed by Drain as “in order to learn the semantics and statistical properties of the natural language” (para [0007]) and determining a difference between vocabulary embeddings is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claim 18, Newman/Reza teach a medium of claim 17.
Newman/Reza do not explicitly teach wherein the first machine-learned model comprises a trained large language model, and the second machine-learned model comprises a trained language model different than the first machine-learned model.
Drain teaches wherein the first machine-learned model comprises a trained large language model, and the second machine-learned model comprises a trained language model different than the first machine-learned model (see Fig. 2, Fig. 8, element 208, element 216, para [0081], discloses two different neural transformer models (first and second machine-learned models)).
Newman/Reza/Drain are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman/Reza to determine a difference between vocabulary embeddings from disclosure of Drain. The motivation to combine these arts is disclosed by Drain as “in order to learn the semantics and statistical properties of the natural language” (para [0007]) and determining a difference between vocabulary embeddings is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claim 19, Newman/Reza teach a medium of claim 17.
Newman/Reza do not explicitly teach wherein determining the difference between the first machine-learned model and the second machine-learned model comprises: determining a difference between vocabulary embeddings of the first machine-learned model and vocabulary embeddings of the second machine-learned model.
Drain teaches wherein determining the difference between the first machine-learned model and the second machine-learned model comprises: determining a difference between vocabulary embeddings of the first machine-learned model and vocabulary embeddings of the second machine-learned model (see Figs. 2-3, Fig. 5, para [0007-0008], para [0034, 0041], discloses determining semantic and statistical differences between a pre-trained neural transformer model and a fine-tuned neural transformer model).
Newman/Reza/Drain are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman/Reza to determine a difference between vocabulary embeddings from disclosure of Drain. The motivation to combine these arts is disclosed by Drain as “in order to learn the semantics and statistical properties of the natural language” (para [0007]) and determining a difference between vocabulary embeddings is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Regarding claim 20, Newman/Reza teach a medium of claim 17.
Newman/Reza do not explicitly teach wherein: determining the difference between the first machine-learned model and the second machine-learned model comprises determining a first linear combination of vocabulary embeddings of the first machine-learned model and the task-specific prompt; and wherein modifying the task-specific prompt comprises determining the updated task- specific prompt based at least in part on the first linear combination and vocabulary embeddings of the second machine-learned model.
Drain teaches wherein: determining the difference between the first machine-learned model and the second machine-learned model comprises determining a first linear combination of vocabulary embeddings of the first machine-learned model and the task-specific prompt (see Fig. 3, Fig. 10, para [0008], para [0086], discloses fine-tuning of neural transformer model, including assert statement prediction and task-specific head of model); and wherein modifying the task-specific prompt comprises determining the updated task- specific prompt based at least in part on the first linear combination and vocabulary embeddings of the second machine-learned model (see Fig. 3, Fig. 10, para [0008], para [0041], discloses fine-tuning neural transformer model with attention with test-assert triplets).
Newman/Reza/Drain are analogous arts as they are each from the same field of endeavor of database systems.
Before the effective filing date of the invention it would have been obvious to a person of ordinary skill in the art to modify the system of Newman/Reza to determine a difference between vocabulary embeddings from disclosure of Drain. The motivation to combine these arts is disclosed by Drain as “in order to learn the semantics and statistical properties of the natural language” (para [0007]) and determining a difference between vocabulary embeddings is well known to persons of ordinary skill in the art, and therefore one of ordinary skill would have good reason to pursue the known options within his or her technical grasp that would lead to anticipated success.
Allowable Subject Matter
Claims 5, 8, 13, and 16 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 following is a statement of reasons for the indication of allowable subject matter: The prior art neither teaches nor suggest the machine-learned prompt recycling model obtaining updated task-specific prompt corresponding to an updated version of a machine-learned model in claims 5 and 13, and prior art does not teach determining a relevance of a first and second subset of a respective first and second vocabulary embeddings and training the machine-learned prompt recycling model based on differences between first and second vocabulary embeddings in claims 8 and 16.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See Selvakumar US Publication No. 2020/0356838.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to COURTNEY HARMON whose telephone number is (571)270-5861. The examiner can normally be reached M-F 9am - 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, Ann Lo can be reached at 571-272-9767. 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.
/Courtney Harmon/Primary Examiner, Art Unit 2159
Prosecution Insights