CTFR 17/506,395 CTFR 92919 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Amendments This action is in response to amendments filed November 24 th , 2025, in which Claims 1 , 8 , and 15 are amended. The amendments have been entered, and Claims 1-20 are currently pending. Claim Rejections - 35 USC § 101 07-04-01 AIA 07-04 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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2016 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”). When considering subject matter eligibility under 35 U.S.C. 101, it must be determined whether the claim is directed to one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter (Step 1). If the claim does fall within one of the statutory categories, the second step in the analysis is to determine whether the claim is directed to a judicial exception (Step 2A). The Step 2A analysis is broken into two prongs. In the first prong (Step 2A, Prong 1), it is determined whether or not the claims recite a judicial exception (e.g., mathematical concepts, mental processes, certain methods of organizing human activity). If it is determined in Step 2A, Prong 1 that the claims recite a judicial exception, the analysis proceeds to the second prong (Step 2A, Prong 2), where it is determined whether or not the claims integrate the judicial exception into a practical application. If it is determined at step 2A, Prong 2 that the claims do not integrate the judicial exception into a practical application, the analysis proceeds to determining whether the claim is a patent-eligible application of the exception (Step 2B). If an abstract idea is present in the claim, any element or combination of elements in the claim must be sufficient to ensure that the claim integrates the judicial exception into a practical application, or else amounts to significantly more than the abstract idea itself. Claim 1 Step 1 : The claim recites a system ; therefore, it is directed to the statutory category of a machine. Step 2A Prong 1 : The claim recites, inter alia: select a subset of deep learning operators and non-deep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a set, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. tune a hyperparameter included in the subset of deep learning and non-deep learning operators, wherein tuning comprises tuning utilizing Gradient-Descent optimization; and : This limitation encompasses the mathematical concept of tuning a parameter with a Gradient-Descent algorithm, which is an evaluation practically capable of being performed in the human mind with the assistance of pen and paper. train the subset of deep learning operators and non-deep learning operators using backpropagation across at least two deep learning operators of the subset of deep learning and non-deep learning operators : This limitation encompasses the mathematical concept of training a subset of at least two deep-learning operators using a backpropagation algorithm, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein in selecting a subset of the deep learning and non-deep learning operators, the system handles both deep learning and non-deep learning operators without implementing separate subsets for the deep learning and non-deep learning operators : This limitation encompasses the mental process of selecting operators from a single subset, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : The additional element, a memory that stores computer executable components , amounts to invoking computers or other machinery merely as a tool to perform an existing process. Thus, this additional element is recited in a manner that represents no more than mere instructions to apply the abstract idea on a computer (see MPEP § 2106.05(f)). The additional element, a processor that executes the computer executable components stored in the memory , amounts to invoking computers or other machinery merely as a tool to perform an existing process. Thus, this additional element is recited in a manner that represents no more than mere instructions to apply the abstract idea on a computer (see MPEP § 2106.05(f)). Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 2 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites: the system of claim 1, wherein the subset of deep learning and non-deep learning operators is selected from a directed acyclic graph comprising deep learning operators and non-deep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a graph, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 3 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites: the system of claim 1, wherein the deep learning operators in the subset of deep learning operators and non-deep learning operators are marked by a higher order operator : This limitation encompasses the mental process of marking higher order operators by identifying those operators which take other operators as input, which is an observation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 4 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites: the system of claim 1, wherein the backpropagation comprises passing of gradients computed from a loss backwards to adjust learned coefficients in the subset of deep learning and non-deep learning operators : This limitation encompasses a mathematical concept (as outlined in the rejection of claim 1) which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 5 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites: the system of claim 1, wherein the subset of deep learning and non-deep learning operators is a specific instantiation of deep learning and non-deep learning operators and hyperparameters : This limitation encompasses the mental process of selecting a static subset of operators and hyperparameters, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein the hyperparameters are tuned automatically : This limitation encompasses a mathematical concept ([0038], “Selection component 104 can tune hyperparameters utilizing Bayesian optimization or [Gradient-Descent] optimization”), which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 6 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites the abstract ideas of the preceding claims from which it depends. Step 2A Prong 2, Step 2B : The additional element, wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in different deep learning frameworks , amounts to no more than generally linking the use of an abstract idea ( the deep learning operators in the subset of deep learning and nondeep learning operators ) to a particular technological environment or field of use ( different deep learning frameworks ) (see MPEP § 2106.05(h)). Claim 7 Step 1 : A machine, as above. Step 2A Prong 1 : The claim recites the abstract ideas of the preceding claims from which it depends. Step 2A Prong 2, Step 2B : The additional element, wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in a same deep learning framework , amounts to no more than generally linking the use of an abstract idea ( the deep learning operators in the subset of deep learning and nondeep learning operators ) to a particular technological environment or field of use ( a same deep learning framework ) (see MPEP § 2106.05(h)). Claim 8 Step 1 : The claim recites a method ; therefore, it is directed to the statutory category of a process. Step 2A Prong 1 : The claim recites, inter alia: selecting…a subset of deep learning operators and nondeep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a set, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. tuning…a hyperparameter included in the subset of deep learning and non-deep learning operators, wherein tuning comprises tuning utilizing Gradient-Descent optimization; and : This limitation encompasses the mathematical concept of tuning a parameter with a Gradient-Descent algorithm, which is an evaluation practically capable of being performed in the human mind with the assistance of pen and paper. The additional element, training…the subset of deep learning operators and non-deep learning operators wherein deep learning operators in the subset of deep learning and non-deep learning operators are trained using backpropagation across at least two deep learning operators of the subset of deep learning and non-deep learning operators : This limitation encompasses the mathematical concept of training a subset of at least two deep-learning operators using a backpropagation algorithm, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein in selecting a subset of the deep learning and non-deep learning operators, the system handles both deep learning and non-deep learning operators without implementing separate subsets for the deep learning and non-deep learning operators : This limitation encompasses the mental process of selecting operators from a single subset, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : The additional elements, by a system operatively coupled to a processor and by the system , amount to invoking computers or other machinery merely as a tool to perform an existing process. Thus, these additional elements are recited in a manner that represents no more than mere instructions to apply the abstract idea on a computer (see MPEP § 2106.05(f)). Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 9 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites: the computer-implemented method of claim 8, wherein the subset of deep learning and non-deep learning operators is selected from a directed acyclic graph comprising deep learning operators and non-deep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a graph, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 10 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites: the computer-implemented method of claim 8, wherein the deep learning operators in the subset of deep learning operators and non-deep learning operators are marked by a higher order operator : This limitation encompasses the mental process of marking higher order operators, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 11 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites: the computer-implemented method of claim 8, wherein the backpropagation comprises passing of gradients computed from a loss backwards to adjust learned coefficients in the subset of deep learning and non-deep learning operators : This limitation encompasses a mathematical concept (as outlined in the rejection of claim 8) which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 12 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites: the computer-implemented method of claim 8, wherein the subset of deep learning and non-deep learning operators is a specific instantiation of deep learning and non-deep learning operators and hyperparameters : This limitation encompasses the mental process of selecting a static subset of operators and hyperparameters, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein the hyperparameters are tuned automatically : This limitation encompasses a mathematical concept ([0038], “Selection component 104 can tune hyperparameters utilizing Bayesian optimization or [Gradient-Descent] optimization”), which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 13 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites the abstract ideas of the preceding claims from which it depends. Step 2A Prong 2, Step 2B : The additional element, wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in different deep learning frameworks , amounts to no more than generally linking the use of an abstract idea ( the deep learning operators in the subset of deep learning and nondeep learning operators ) to a particular technological environment or field of use ( different deep learning frameworks ) (see MPEP § 2106.05(h)). Claim 14 Step 1 : A process, as above. Step 2A Prong 1 : The claim recites the abstract ideas of the preceding claims from which it depends. Step 2A Prong 2, Step 2B : The additional element, wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in a same deep learning framework , amounts to no more than generally linking the use of an abstract idea ( the deep learning operators in the subset of deep learning and nondeep learning operators ) to a particular technological environment or field of use ( a same deep learning framework ) (see MPEP § 2106.05(h)). Claim 15 Step 1 : The claim recites a computer program product ; therefore, it is directed to the statutory category of an article of manufacture ([0080], “The term article of manufacture, as used herein, is intended to encompass a computer program accessible from any computer-readable device or storage media”). Step 2A Prong 1 : The claim recites, inter alia: select…a subset of deep learning operators and nondeep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a set, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. tune…a hyperparameter included in the subset of deep learning and non-deep learning operators, wherein tuning comprises tuning utilizing Gradient-Descent optimization; and : This limitation encompasses the mathematical concept of tuning a parameter with a Gradient-Descent algorithm, which is an evaluation practically capable of being performed in the human mind with the assistance of pen and paper. train…the subset of deep learning operators and non-deep learning operators wherein deep learning operators in the subset of deep learning and non-deep learning operators are trained using backpropagation across at least two deep learning operators of the subset of deep learning and non-deep learning operators : This limitation encompasses the mathematical concept of training a subset of at least two deep-learning operators using a backpropagation algorithm, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein in selecting a subset of the deep learning and non-deep learning operators, the system handles both deep learning and non-deep learning operators without implementing separate subsets for the deep learning and non-deep learning operators : This limitation encompasses the mental process of selecting operators from a single subset, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : The additional element, the computer program product comprising one or more computer readable storage medium having program instructions imbodied therewith , amounts to invoking computers or other machinery merely as a tool to perform an existing process. Thus, this additional element is recited in a manner that represents no more than mere instructions to apply the abstract idea on a computer (see MPEP § 2106.05(f)). The additional elements, the program instructions executable by a processor and by the processor , amount to invoking computers or other machinery merely as a tool to perform an existing process. Thus, these additional elements are recited in a manner that represents no more than mere instructions to apply the abstract idea on a computer (see MPEP § 2106.05(f)). Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 16 Step 1 : An article of manufacture, as above. Step 2A Prong 1 : The claim recites: the computer program product of claim 15, wherein the subset of deep learning and non-deep learning operators is selected from a directed acyclic graph comprising deep learning operators and non-deep learning operators : This limitation encompasses the mental process of selecting a subset of operators from a graph, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 17 Step 1 : An article of manufacture, as above. Step 2A Prong 1 : The claim recites: the computer program product of claim 15, wherein the deep learning operators in the subset of deep learning operators and non-deep learning operators are marked by a higher order operator : This limitation encompasses the mental process of marking higher order operators, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 18 Step 1 : An article of manufacture, as above. Step 2A Prong 1 : The claim recites: the computer program product of claim 15, wherein the backpropagation comprises passing of gradients computed from a loss backwards to adjust learned coefficients in the subset of deep learning and non-deep learning operators : This limitation encompasses a mathematical concept (as outlined in the rejection of claim 15) which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 19 Step 1 : An article of manufacture, as above. Step 2A Prong 1 : The claim recites: the computer program product of claim 15, wherein the subset of deep learning and non-deep learning operators is a specific instantiation of deep learning and non-deep learning operators and hyperparameters : This limitation encompasses the mental process of selecting a static subset of operators and hyperparameters, which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. wherein the hyperparameters are tuned automatically : This limitation encompasses a mathematical concept ([0038], “Selection component 104 can tune hyperparameters utilizing Bayesian optimization or [Gradient-Descent] optimization”), which is an evaluation that is practically capable of being performed in the human mind with the assistance of pen and paper. Step 2A Prong 2, Step 2B : Nothing in the claim integrates the abstract idea into a practical application, nor does the claim include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception. Therefore, the claim is subject-matter ineligible. Claim 20 Step 1 : An article of manufacture, as above. Step 2A Prong 1 : The claim recites the abstract ideas of the preceding claims from which it depends. Step 2A Prong 2, Step 2B : The additional element, wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in different deep learning frameworks , amounts to no more than generally linking the use of an abstract idea ( the deep learning operators in the subset of deep learning and nondeep learning operators ) to a particular technological environment or field of use ( different deep learning frameworks ) (see MPEP § 2106.05(h)). Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA 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. 07-21-aia AIA Claim s 1 , 2, 4-6, 8, 9, 11-13, 15, 16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Shen et al. (US 20190228297 A1, hereinafter Shen) in view of Milutinovic et al. (“End-to-end Training of Differentiable Pipelines Across Machine Learning Frameworks,” hereinafter Milutinovic) in further view of Pedregosa et al., “Hyperparameter optimization with approximate gradient.” Regarding claim 1 , Shen teaches a system comprising: a memory and (Fig. 6, elements 630 and 640; “Memory” and “Storage”; [0071], “Computer readable storage media includes…media implemented in any method or technology for storage of information such as computer-readable instructions…Memory 630 and Storage 640 are all examples of computer-readable storage media.”) a processor coupled to the memory, wherein the processor is configured to (Shen Fig. 6, element 620; “Processing Unit”; Shen [0070]; “Computing Device 610 typically includes at least one Central Processing Unit (CPU) 620 and Memory 630.”) select a subset of deep learning and non-deep learning operators; and (Shen [0051]; “Operators may be deep learning operators or non-deep learning.” Shen [0024]; “With a graph filled with many possible operators…Artificial Intelligence Modeling Engine 200 may determine which operators to use” thereby select[ing] a subset . ) tune a hyperparameter included in the subset of deep learning and non-deep learning operators, wherein tuning comprises tuning utilizing … optimization; and (Shen, [0057]; “Meta-learning may use machine learning models to predict hyperparameters” Shen, [0062]; “This optimization problem may be addressed using machine learning approaches” wherein “to predict hyperparameters” as part of an “optimization problem” ) train the subset of deep learning operators and non-deep learning operators (Shen [0020-21]; “Operators may also be layers or multiple layers of neural operators…Neural network layers may communicate both ways, rather than taking one or more inputs and generating outputs. They hold internal states or weights, and the weights may be updated by the operators that follow after them through back-propagation…Training may include passing training data through an operator and providing correct outputs to the operator for the training data, which may allow the operator to update internal weighting or other status so that the operator may provide correct outputs for other data,” wherein the “back-propagation” and the “Training…to update internal weighting” encompasses the function of the training component outlined in the specification of the present invention at [0042], “training component 105 can train two or more deep learning operators using backpropagation by repeatedly executing three phases: a forward pass, a backwards pass, and updating learned coefficients using a stochastic gradient decent (SGD) optimizer.”). Shen does not explicitly teach using backpropagation across at least two deep learning operators of the subset of deep learning and non-deep learning operators . However, Milutinovic, in the area of training deep learning pipelines using backpropagation, teaches this limitation (Milutinovic, 2 End-to-End Optimization of Machine Learning Pipelines, pp. 2, paragraph 4; “Note that primitives,” which in this context are equivalent to operators , “can encapsulate computations of varying levels of granularity, ranging from simple-dimensionality-reduction transformations to full-scale image- or speech-recognition models,” thereby encompassing deep learning operators in accordance with the definition provided in the specification of the current invention at [0037] , “In a scikit-learn style pipeline, deep learning operators can behave like transformers, first or intermediate operators in a pipeline, or like estimators, final operators in a pipeline. Transformers can support a transform method, whereas estimators can support regression or classification with class probabilities via a predict_proba method. In neural networks, first or intermediate layers of the network can be represented as deep learning featurizer operators and final layers of the network can be represented as deep learning head operators.” Milutinovic, 2 End-to-End Optimization of Machine Learning Pipelines, pp. 2, paragraph 5; “A differentiable pipeline is one which can be differentiated end-to-end, requiring that there is at least one path from x to y through which every primitive can compute at least the gradient of its outputs with respect to its inputs, allowing backpropagation from y to x ,” an example of which is given at Milutinovic, 3 Example Pipeline, pp. 4, paragraph 2; “By each component providing a backprop method, we are able to jointly train the classifier, the dimensionality reduction, and the neural network, also fine-tuning the weights of the pre-trained CNN in the process,” wherein “the classifier” is a convolutional neural network and “the neural network” refers to a multiple linear regression model implemented as a neural network; therefore, backpropagation is used for training across at least two deep learning operators ) Milutinovic is analogous to the claimed invention as both are from the same field of endeavor, that is, inter-operator backpropagation in machine learning pipelines. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the subset selection and training steps of Shen with the differential deep-learning pipeline of Milutinovic. The motivation to do so is to facilitate backpropagation through an entire machine learning pipeline, rather than training each individual operator or model individually, as suggested by Milutinovic (Milutinovic, 2 End-to-End Optimization of Machine Learning Pipelines, pp. 3, paragraph 6; “This formulation,” referring the method of backpropagation cited above, “allows the loss on end-to-end data to be calculated and backpropagated through the pipeline. We propose that primitives combine this with the contributions to the gradient from the parameter priors and local training data in their backprop method,” the benefits of which are shown in the previously stated example at Milutinovic, 3 Example Pipeline, pp. 4, paragraph 2; “By each component providing a backprop method, we are able to jointly train the classifier, the dimensionality reduction, and the neural network” (emphasis added).). Shen further teaches wherein in selecting a subset of the deep learning and non-deep learning operators, the system handles both deep learning and non-deep learning operators without implementing separate subsets for the deep learning and non-deep learning operators, (Shen [0051]; “Operators may be deep learning operators or non-deep learning. A deep learning operator may be one which learns data representations, while a non-deep operator may learn task-specific algorithms. A linear regression operator, for example, may be a non-deep learning operator. In contrast, a convolutional neural net (CNN) operator may be a deep learning operator, and may be parameterized by stripe, filters, activation, or other attributes.” The broadest reasonable interpretation of “operators may be deep learning operators of non-deep learning” encompasses the inclusive subset containing both deep learning and non-deep learning operators . Therefore, deep learning and non-deep learning operators are not in separate subsets . ) Shen teaches performing hyperparameter tuning, but does not teach to do so using gradient-descent optimization. However, Pedregosa teaches hyperparameter tuning via gradient-descent optimization (Pedregosa, title, “Hyperparameter optimization with approximate gradient” & Fig. 1, “to estimate the optimal hyperparameters by gradient descent”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Shen to use a gradient descent optimization for tuning hyperparameters, rather than Shen’s optimization. The motivation to do so is that “our approach [gradient descent] is highly competitive with respect to state of the art methods” (Pedregosa, Abstract & pg. 8, Fig. 3, showing that the method of Pedregosa is often superior to SMBO, i.e. Bayesian Optimization ). Regarding claim 2 , the combination of Shen, Milutinovic and Pedregosa teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen further teaches wherein the subset of deep learning and non-deep learning operators is selected from a directed acyclic graph comprising deep learning operators and non-deep learning operators (Shen [0024], “With a graph filled with many possible operators, a user may specify some aspects of what the type of data the user has and what the user may want to have predicted, and Artificial Intelligence Modeling Engine 200 may determine which operators to use.” Shen Figs. 2 and 3, both figures specify the graph as a directed acyclic graph as each edge possesses a defined direction and no loops are possible .). Regarding claim 4 , the combination of Shen, Milutinovic and Pedregosa teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen does not explicitly teach wherein the backpropagation comprises passing of gradients computed from a loss backwards to adjust learned coefficients in the subset of deep learning and non-deep learning operators . However, Milutinovic, in the area of training deep learning pipelines using backpropagation, teaches this limitation (Milutinovic, 2 End-to-End Optimization of Machine Learning Pipelines, pp. 3, paragraph 6; “This formulation,” referring the method of backpropagation cited in the rejection of claim 1, “allows the loss on end-to-end data to be calculated and backpropagated through the pipeline. We propose that primitives combine this with the contributions to the gradient from the parameter priors,” which are equivalent to the learned coefficients , “and local training data in their backprop method,” this method denoting an adjust[ment] of the gradients . ). Milutinovic is analogous to the claimed invention as both are from the same field of endeavor, that is, inter-operator backpropagation in machine learning pipelines. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the subset selection and training steps of Shen with the differential deep-learning pipeline of Milutinovic. The motivation to do so is to facilitate backpropagation through an entire machine learning pipeline, rather than training each individual operator or model individually, as suggested by Milutinovic (Milutinovic, 2End-to-End Optimization of Machine Learning Pipelines, pp. 3, paragraph 6; “This formulation,” referring the method of backpropagation cited above, “allows the loss on end-to-end data to be calculated and backpropagated through the pipeline. We propose that primitives combine this with the contributions to the gradient from the parameter priors and local training data in their backprop method,” the benefits of which are shown in the previously stated example at Milutinovic pg. 4, Example Pipeline; “By each component providing a backprop method, we are able to jointly train the classifier, the dimensionality reduction, and the neural network” (emphasis added).). Regarding claim 5 , the combination of Shen, Milutinovic and Pedregosa teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen further teaches wherein the subset of deep learning and non-deep learning operators is a specific instantiation of deep learning and non-deep learning operators and hyperparameters (Shen [0046]; “Given a meta-learning graph, Artificial Intelligence Modeling Engine 200 may generate instantiated graphs for a specific dataset. This may be done using select_n and hyperparameter operators.” Shen [0037]; “a hyperparameter operator may estimate a hyperparameter K…”) wherein the hyperparameters are tuned automatically (Shen [0056-57]; “Hyperparameters may be estimated using machine learning models. Meta-learning may use machine learning models to predict hyperparameters.” [T]he hyperparameters are “estimated” or “predicted” (e.g., tuned ) automatically by “machine learning models,” rather than manually by the user. ). Regarding claim 6 , the combination of Shen, Milutinovic and Pedregosa teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen does not explicitly teach wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in different deep learning frameworks . However, Milutinovic, in the area of training deep learning pipelines using backpropagation, teaches this limitation (Milutinovic, 3 Example Pipeline, pp. 4, paragraph 2; “Every primitive of this pipeline is written in a different machine learning framework. The CNN loads a pre-trained model published in Model Zoo and uses Caffe. The PCA primitive wraps the PCA implementation in scikit-learn. The linear regression primitive is written in PyTorch”). Milutinovic is analogous to the claimed invention as both are from the same field of endeavor, that is, inter-operator backpropagation in machine learning pipelines. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to implement the machine learning pipeline of Shen, Milutinovic and Pedregosa in a plurality of frameworks, as further taught by Milutinovic. The motivation to do so is to utilize each framework’s supply of pre-trained deep learning operators thereby cutting down on training time and reducing development and resource costs (Milutinovic, Abstract; “This is distinguished from recent interoperability efforts such as the Open Neural Network Exchange (ONNX) format and other language-centric cross compilation approaches in that the final pipeline does not need to be implemented nor trained in the same language nor cross-compiled into any single language; in other words. primitives may be written and pre-trained in PyTorch, TensorFlow. Caffe, scikit-learn or any of the other popular machine learning frameworks and fine-tuned end-to-end while being executed directly in their host frameworks.”). Claims 8, 9 and 11-13 are method claims corresponding to the steps of claims 1, 2 and 4-6 and are thus rejected for the same reasons. Claims 15, 16 and 18-20 are product claims corresponding to the steps of claims 1, 2 and 4-6 and are thus rejected for the same reasons . 07-21-aia AIA Claim s 3, 7, 10, 14, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Shen in view of Milutinovic, Pedregosa and Hu et al. (“Jittor: a novel deep learning framework with meta-operators and unified graph execution,” hereinafter Hu) . Regarding claim 3 , the combination of Shen, Milutinovic and Pedregosa teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen does not explicitly teach wherein the deep learning operators in the subset of deep learning operators and non-deep learning operators are marked by a higher order operator . However, Hu, in the area of deep learning pipelines, teaches this limitation (Hu, 3.1 Meta-operators, pp. 4-5, paragraphs 8 and 1; “A meta-operator is a general operator, which when specialized gives a class of operators with common properties that make them particularly amenable to optimization… [An] important meta-operator is reindex-reduce, which provides a many-to-one mapping. Sum and product are particular examples of reindex-reduce operators,” wherein “sum and product” are examples of non-deep learning operators . “These meta-operator classes are shown in Figure 2, which also shows how Jittor provides common, higher-level, deep learning operators (e.g., convolution, normalization, and pooling), by fusing meta-operators” thus detailing how the combination of the non-deep learning operators from the “meta-operator classes” forms “higher-level” or higher order operator[s] in the manner defined in the specification of the present invention at [0041] , “deep learning operators in the subset of deep learning and non-deep learning operators can be marked by a higher order, or meta-order, operator. A higher order operator can be defined as an operator that takes other operators as input.” Hu Figure 2; the figure illustrates how the fusion of the non-deep learning operators from the “meta-operator classes” creates higher order deep learning operators ). Hu is analogous to the claimed invention as both are from the same field of endeavor, that is, machine learning pipelines. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the subset selection and training method of Shen, Milutinovic and Pedregosa with the meta-operators of Hu. The motivation to do so is to use combinations of non-deep learning operators to create deep learning operators as taught by Hu (Hu, 3.2 Fusion of operators, pp. 6, paragraph 6; “Listing 1 shows how a convolution operator can be implemented in terms of meta-operators. A general reindex meta-operator is used, in addition to specialized broadcast and sum operators.” Hu Listing 1; “Python implementation of convolution using three operators: reindex, broadcast, and sum.”). Regarding claim 7 , the combination of Shen, Milutinovic, Pedregosa and Hu teaches the system of claim 1 (and thus the rejection of claim 1 is incorporated). Shen does not explicitly teach wherein the deep learning operators in the subset of deep learning and non-deep learning operators are implemented in a same deep learning framework . However, Hu, in the area of deep learning pipelines, teaches this limitation (Hu, 3.1 Meta-operators, pp. 4-5; “These meta-operator classes are shown in Figure 2, which also shows how Jittor provides common, higher-level, deep learning operators (e.g., convolution, normalization, and pooling), by fusing meta-operators,” wherein “Jittor” is the same deep learning framework in which the listed deep learning operators will be implemented .). Hu is analogous to the claimed invention as both are from the same field of endeavor, that is, machine learning pipelines. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to implement the machine learning pipeline of Shen, Milutinovic and Pedregosa in the Jittor framework disclosed by Hu. The motivation to do so is to use Jittor’s fusion functionality to implement a pipeline of non-deep learning and deep learning operators (Hu, 3.2 Fusion of operators, pp. 6, paragraph 6; “Listing 1 shows how a convolution operator can be implemented in terms of meta-operators. A general reindex meta-operator is used, in addition to specialized broadcast and sum operators.” Hu Listing 1; “Python implementation of convolution using three operators: reindex, broadcast, and sum.”). Claims 10 and 14 are method claims corresponding to the steps of claims 3 and 7 and are thus rejected for the same reasons. Claim 17 is a product claim corresponding to the steps of claim 3 and is thus rejected for the same reasons. Response to Arguments Applicant’s arguments filed November 24 th , 2025 have been fully considered, but are not fully persuasive. Applicant’s arguments regarding the 35 U.S.C. 101 rejection as directed towards an abstract idea have been fully considered, but are not persuasive. Applicant argues that “The claimed invention addresses reduction in the workload of the processor … inter-operator system 101 can handle deep learning and non-deep learning operations without implementing separate subsets”. However, applicant fails to recite any steps that actually achieve this capability. Applicant is merely stating the result or outcome, where (as per MPEP 2106.05(f)(1)) “the claim fails to recite details of how a solution to a problem is accomplished”. The independent claims are analogous to the inventor of a paint-roller claiming merely “A method of painting a house, comprising: applying paint to the house without using a paint brush” – the claim fails to recite any specific steps which achieve the result of being able to handle deep learning and non-deep learning operations together. By the analysis of MPEP 2106.05(f)(2), the claims as recited are subject-matter ineligible. Applicant’s arguments regarding the 35 U.S.C. 103 rejection of the independent claims as been fully considered, but is 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. Specifically, new reference Pedregosa teaches hyperparameters optimization using gradient descent methods rather than Bayesian optimization methods. Applicant’s arguments regarding the dependent claims rely upon features argued with respect to the independent claims, and are similarly unpersuasive. Conclusion 07-40 AIA 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 BRIAN M SMITH whose telephone number is (469)295-9104. The examiner can normally be reached Monday - Friday, 8:00am - 4pm Pacific. 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, Kakali Chaki can be reached at (571) 272-3719. 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. /BRIAN M SMITH/Primary Examiner, Art Unit 2122 Application/Control Number: 17/506,395 Page 2 Art Unit: 2122 Application/Control Number: 17/506,395 Page 3 Art Unit: 2122 Application/Control Number: 17/506,395 Page 4 Art Unit: 2122 Application/Control Number: 17/506,395 Page 5 Art Unit: 2122 Application/Control Number: 17/506,395 Page 6 Art Unit: 2122 Application/Control Number: 17/506,395 Page 7 Art Unit: 2122 Application/Control Number: 17/506,395 Page 8 Art Unit: 2122 Application/Control Number: 17/506,395 Page 9 Art Unit: 2122 Application/Control Number: 17/506,395 Page 10 Art Unit: 2122 Application/Control Number: 17/506,395 Page 11 Art Unit: 2122 Application/Control Number: 17/506,395 Page 12 Art Unit: 2122 Application/Control Number: 17/506,395 Page 13 Art Unit: 2122 Application/Control Number: 17/506,395 Page 14 Art Unit: 2122 Application/Control Number: 17/506,395 Page 15 Art Unit: 2122 Application/Control Number: 17/506,395 Page 16 Art Unit: 2122 Application/Control Number: 17/506,395 Page 17 Art Unit: 2122 Application/Control Number: 17/506,395 Page 18 Art Unit: 2122 Application/Control Number: 17/506,395 Page 19 Art Unit: 2122 Application/Control Number: 17/506,395 Page 20 Art Unit: 2122 Application/Control Number: 17/506,395 Page 21 Art Unit: 2122 Application/Control Number: 17/506,395 Page 22 Art Unit: 2122 Application/Control Number: 17/506,395 Page 23 Art Unit: 2122 Application/Control Number: 17/506,395 Page 24 Art Unit: 2122 Application/Control Number: 17/506,395 Page 25 Art Unit: 2122 Application/Control Number: 17/506,395 Page 26 Art Unit: 2122 Application/Control Number: 17/506,395 Page 27 Art Unit: 2122 Application/Control Number: 17/506,395 Page 28 Art Unit: 2122
Prosecution Insights