CTNF 18/415,261 CTNF 73675 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. This office action is in the claimed invention filed on January 17, 2024, in which claims 1-20 are presented for examination. Information Disclosure Statement The information disclosure statement filed on June 21, 2024, April 18, 2025 and March 10, 2026 and June 08, 2026 complies with the provisions of 37 CFR 1.97, 1.98 and MPEP § 609. It has been placed in the application file. The information referred to therein has been considered as to the merits. 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 without significantly more. Step 1, Statutory Category : Claims 1-9 are directed to a method Claims 10-18 are directed to an electronic device Claims 19-20 are directed to a non-transitory computer readable storage medium. Therefore, claims 1-20 fall into at least one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter. Step 2A, Prong One (Judicial exception recited) The limitation “identifying, by the electronic device, an optimal Pareto front from among the plurality of Pareto fronts” in claims 1, 10 and 19, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. That is, other than reciting “identifying, by the electronic device, an optimal Pareto front from among the plurality of Pareto fronts”, nothing in the claim element precludes the steps from practically being performed in a human mind. For example, the limitation “identifying, by the electronic device, an optimal Pareto front from among the plurality of Pareto fronts”, in the context of these claims encompasses one can mentally, or manually with the aid of pen and paper identifies an optimal Pareto front from among the plurality of Pareto fronts. The limitation “identifying, by the electronic device, the at least two performance parameters corresponding to the second AI model based on identifying that the second AI model belongs to the optimal Pareto front” in claims 1, 10 and 19, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. That is nothing in the claim element precludes the steps from practically being performed in a human mind. For example, the language, “identifying, by the electronic device, the at least two performance parameters corresponding to the second AI model based on identifying that the second AI model belongs to the optimal Pareto front”, in the context of the claim encompasses one can manually with the aid of pen and paper identifies the at least two performance parameters corresponding to the second AI model based on identifying that the second AI model belongs to the optimal Pareto front. If a claim limitation, under its broadest reasonable interpretation, covers mental processes but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas (concepts performed in the human mind including an observation, evaluation, judgement, and opinion). Accordingly, the claim recites an abstract idea. Step 2A, Prong Two ( Integrated into a practical application): This judicial exception is not integrated into a practical application. In particular, the claims recite the following additional elements: The limitation “providing, by the electronic device, a plurality of Pareto fronts based on at least two performance parameters” amounts to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)). The limitation “providing, by the electronic device, a second AI model iteratively; identifying, by the electronic device, whether the second AI model belongs to the optimal Pareto front” amounts to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)). The limitation “obtaining, by the electronic device, the second AI model based on identifying that the second AI model meets one or more predetermined performance parameters” amounts to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)). The limitation “a memory, communication, at least a processor, and non-transitory computer-readable storage medium” are recited at a high level of generality such that they amount to on more than mere instructions to apply the exception using a generic component. (see MPEP 2106.05(f)). These limitations can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of a computer (see MPEP 2106.05(h)). Note, the mere instructions to apply an exception on a generic computer cannot integrate a judicial exception into a practical application. Step 2B ( claim provides an inventive concept): The conclusions for the mere implementation using a computer, mere field of use, and using generic computer components (Artificial intelligent i.e. AI) as a tool are carried over and do not provide significantly more. With respect to the "providing …..” identified as insignificant extra-solution activity above when re-evaluated this element is well-understood, routine, and conventional as evidenced by the court cases in MPEP 2106.05(d)(II), "iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334; i. … transmitting data over a network, …Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); … OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network)". With respect to the “obtaining …" identified as insignificant extra-solution activity above when re-evaluated this element is well-understood, routine, and conventional as evidenced by the court cases in MPEP 2106.05(d)(II), "i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); … OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network);" and thus remains insignificant extra-solution activity that does not provide significantly more. With respect to the “a memory, communication, at least a processor, and non-transitory computer-readable storage medium” amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015). The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements when considered both individually and as an ordered combination do not amount to significantly more than the abstract idea. Looking at the claim as a whole does not change this conclusion and the claim appears to be ineligible. Accordingly, claim 1 is directed to an abstract idea. The remaining independent claim 10 and 19 fall short the 35 USC 101 requirement under the same rationale. The dependent claims 2-9, 11-18 and 20 when analyzed and each taken as a whole are held to be patent ineligible under 35 USC 101 because the additional recited limitations fail to establish that the claims are not directed to an abstract idea. Claim 2 recites “providing, by the electronic device, a plurality of first AI models”. This additional element is recited at a high level of generality and would function in its ordinary capacity for providing, by the electronic device, a plurality of first AI models, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more. “identifying, by the electronic device, the at least two performance parameters corresponding to each first AI model from among the plurality of first AI models”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claims 11 and 20. Claim 3 recites “wherein the plurality of first AI models comprises at least one of a random AI model or a set of predetermined AI designs”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 12. Claim 4 recites “wherein the optimal Pareto front comprises a plurality of AI models which are equally significant with respect to a plurality of objectives”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 13. Claim 5 recites “wherein the optimal Pareto front dominates the plurality of Pareto fronts with respect to a plurality of objectives”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 14. Claim 6 recites “wherein the identifying whether the second AI model belongs to the optimal Pareto front comprises using a classifier”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 15. Claim 7 recites “wherein the at least two performance parameters are determined based on a plurality of objectives”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 16. Claim 8 recites “wherein the classifier is a binary classifier”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 17. Claim 9 recites “wherein the classifier identifies the optimal Pareto front by identifying each probability that the second AI models belongs to the plurality of Pareto front”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. Same rationale applies to claim 18. Claim Rejections - 35 USC § 103 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 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al., (hereinafter “Ma”), article entitled “How to Simplify Search: Classification-wise Pareto Evolution for One-shot Neural Architecture Search” in view of Chen et al., (hereinafter “Chen”) CN-112257840 . As to claim 1, Ma discloses a method for providing a neural architecture search (NAS) in an electronic device (abstract, deployment of deep neural models, how to effectively and automatically find feasible deep models under diverse design objectives is fundamental. Most existing neural architecture search (NAS) methods utilize surrogates to predict the detailed performance (e.g., accuracy and model size) of a candidate architecture during the search, which however is complicated and inefficient. In contrast, we aim to learn an efficient Pareto classifier to simplify the search process of NAS by transforming the complex multi-objective NAS task into a simple Pareto-dominance classification task), wherein the method comprises: providing, by the electronic device, a plurality of Pareto fronts based on at least two performance parameters (see abstract, propose a classification wise Pareto evolution approach for one-shot NAS, where an online classifier is trained to predict the dominance relationship between the candidate and constructed reference architectures, instead of using surrogates to fit the objective functions); identifying, by the electronic device, an optimal Pareto front from among the plurality of Pareto fronts (page 1, col.1, N EURAL architecture search (NAS) is shown to be promising for the automatic design of task-specific deep neural networks (DNNs), instead of the traditional manual design based on extensive human expertise [1], [2], [3], [4]. Consequently, NAS has received a surge of attention from the community of deep learning, largely owing to its superiority in optimizing the architecture and weights of DNN [2]. In other words, NAS can obtain desirable architectures as human experts do, or even far more innovative architectures [3]. Early NAS works in a nested manner, where the architecture is trained from scratch. Then, the called one-short NAS uses weight sharing to accelerate the training of the architecture [4], [5]. In one-shot NAS, the search space is encoded as a supernet, and the weights of all possible architectures (or subnets) directly inherit from the supernet, without being trained from scratch, while the candidate architectures are evaluated on the validation set [6], [7], [8]. For reaching the optima of the objectives of NAS such as the accuracy and model size (#params), different optimizers can be used in NAS, including evolution algorithms (EAs), reinforcement learning (RL), and gradient methods. EA-based NAS). Ma does not disclose the claimed: providing, by the electronic device, a second AI model iteratively; identifying, by the electronic device, whether the second AI model belongs to the optimal Pareto front; identifying, by the electronic device, the at least two performance parameters corresponding to the second AI model based on identifying that the second AI model belongs to the optimal Pareto front; and obtaining, by the electronic device, the second AI model based on identifying that the second AI model meets one or more predetermined performance parameters. Meanwhile, Chen disclose the claimed “providing, by the electronic device, a second AI model iteratively (page 4, par. [2]-[3], , the feedback function is pre-set with a first performance parameter and a first threshold value, the first performance parameter is any index of the performance index in addition to the accuracy of the neural network, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server according to the accuracy of the third neural network; the second performance parameter of the third neural network and the feedback function to generate the feedback value, wherein the difference between the second performance parameter and the first performance parameter is less than the first threshold value, the feedback value is positive number, the relation between the accuracy of the third neural network and the feedback value is positive correlation; the relationship between the difference between the second performance parameter and the first performance parameter and the feedback value is negative correlation; under the condition that the difference value of the second performance parameter and the first performance parameter is greater than the first threshold value, the value of the feedback value is zero or negative number; the server according to the performance information of the third neural network and feedback function to generate feedback value, the method further comprises: The server increases the first performance parameter. In one possible implementation of the first aspect, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance); identifying, by the electronic device, whether the second AI model belongs to the optimal Pareto front (see page 4, par [3], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance); identifying, by the electronic device, the at least two performance parameters corresponding to the second AI model based on identifying that the second AI model belongs to the optimal Pareto front (see page 4, par. [1]-[2], providing a specific implementation process of generating a third neural network, improving the implementation of the solution; In addition, because the temperature parameter T in the normalized index function is larger, the probability of each first neural unit in the first circulating neural network sampling to L first neural units tends to average, then the third neural network of the server sampling is random. and the temperature parameter T is larger; the server will generate the third neural network according to the strategy learned by the first circulating neural network, and T is changed according to the rule of the cosine function, which can avoid the parameter of the first circulating neural network to converge to a local optimal value and not jump out; that is to ensure the last N second neural network is selected from a large number of the first neural network, and not selected from the local small first neural network, so as to ensure the performance of the N second neural network. In one possible implementation of the first aspect, the feedback function is pre-set with a first performance parameter and a first threshold value, the first performance parameter is any index of the performance index in addition to the accuracy of the neural network, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server according to the accuracy of the third neural network; the second performance parameter of the third neural network and the feedback function to generate the feedback value, wherein the difference between the second performance parameter and the first performance parameter is less than the first threshold value, the feedback value is positive number, the relation between the accuracy of the third neural network and the feedback value is positive correlation; the relationship between the difference between the second performance parameter and the first performance parameter and the feedback value is negative correlation; under the condition that the difference value of the second performance parameter and the first performance parameter is greater than the first threshold value, the value of the feedback value is zero or negative number; the server according to the performance information of the third neural network and feedback function to generate feedback value, the method further comprises: The server increases the first performance parameter); and obtaining, by the electronic device, the second AI model based on identifying that the second AI model meets one or more predetermined performance parameters (see page 4, par. [1]-[2], providing a specific implementation process of generating a third neural network, improving the implementation of the solution; In addition, because the temperature parameter T in the normalized index function is larger, the probability of each first neural unit in the first circulating neural network sampling to L first neural units tends to average, then the third neural network of the server sampling is random. and the temperature parameter T is larger; the server will generate the third neural network according to the strategy learned by the first circulating neural network, and T is changed according to the rule of the cosine function, which can avoid the parameter of the first circulating neural network to converge to a local optimal value and not jump out; that is to ensure the last N second neural network is selected from a large number of the first neural network, and not selected from the local small first neural network, so as to ensure the performance of the N second neural network. In one possible implementation of the first aspect, the feedback function is pre-set with a first performance parameter and a first threshold value, the first performance parameter is any index of the performance index in addition to the accuracy of the neural network, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server according to the accuracy of the third neural network; the second performance parameter of the third neural network and the feedback function to generate the feedback value, wherein the difference between the second performance parameter and the first performance parameter is less than the first threshold value, the feedback value is positive number, the relation between the accuracy of the third neural network and the feedback value is positive correlation; the relationship between the difference between the second performance parameter and the first performance parameter and the feedback value is negative correlation; under the condition that the difference value of the second performance parameter and the first performance parameter is greater than the first threshold value, the value of the feedback value is zero or negative number; the server according to the performance information of the third neural network and feedback function to generate feedback value, the method further comprises: The server increases the first performance parameter). Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to have modified the system od Ma with the system od Chen in order to perform iterative training on the at least one fifth neural network selected by the client, then at least one fifth neural network finishing the iterative training operation is sent to the client, so that the iterative training operation executed by the server is more pertinent, and the waste of computer resource is avoided, thereby ensuring the performance of the N second neural networks. As to claim 2, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein before the providing the plurality of Pareto fronts based on the at least two performance parameters further comprises: providing, by the electronic device, a plurality of first AI models; and identifying, by the electronic device, the at least two performance parameters corresponding to each first AI model from among the plurality of first AI models (see page 4, par. [2], the feedback function is pre-set with a first performance parameter and a first threshold value, the first performance parameter is any index of the performance index in addition to the accuracy of the neural network, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server according to the accuracy of the third neural network; the second performance parameter of the third neural network and the feedback function to generate the feedback value, wherein the difference between the second performance parameter and the first performance parameter is less than the first threshold value, the feedback value is positive number, the relation between the accuracy of the third neural network and the feedback value is positive correlation; the relationship between the difference between the second performance parameter and the first performance parameter and the feedback value is negative correlation; under the condition that the difference value of the second performance parameter and the first performance parameter is greater than the first threshold value, the value of the feedback value is zero or negative number; the server according to the performance information of the third neural network and feedback function to generate feedback value, the method further comprises: The server increases the first performance parameter). As to claim 3, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the plurality of first AI models comprises at least one of a random AI model or a set of predetermined AI designs (see page 4, par. [2], the feedback function is pre-set with a first performance parameter and a first threshold value, the first performance parameter is any index of the performance index in addition to the accuracy of the neural network, the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server according to the accuracy of the third neural network; the second performance parameter of the third neural network and the feedback function to generate the feedback value, wherein the difference between the second performance parameter and the first performance parameter is less than the first threshold value, the feedback value is positive number, the relation between the accuracy of the third neural network and the feedback value is positive correlation; the relationship between the difference between the second performance parameter and the first performance parameter and the feedback value is negative correlation; under the condition that the difference value of the second performance parameter and the first performance parameter is greater than the first threshold value, the value of the feedback value is zero or negative number; the server according to the performance information of the third neural network and feedback function to generate feedback value, the method further comprises: The server increases the first performance parameter). As to claim 4, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the optimal Pareto front comprises a plurality of AI models which are equally significant with respect to a plurality of objectives (see page 4, par. [3]-[4], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance. In one possible implementation of the first aspect, the method may further include: the server receives the neural network selection information sent by the client, wherein the neural network selection information can carry the identification information of each fifth neural network in at least one fifth neural network, also can carry the function corresponding to each fifth neural network; so that the server can select at least one fifth neural network from the N second neural network according to the neural network selection information, then performing iterative training for each fifth neural network, until the convergence condition of the loss function is satisfied, at least one fifth neural network finishing the iterative training operation is sent to the client. In the implementation manner, the server after sending the N second neural network to the client, the client can select at least one fifth neural network from the N second neural network, and then the server performs iterative training for at least one fifth neural network selected by the client; then the iteration training operation of at least one fifth neural network is sent to the client, so that the server performs iterative training operation more pertinence, avoiding the waste of the computer resource). As to claim 5, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the optimal Pareto front dominates the plurality of Pareto fronts with respect to a plurality of objectives (see page 4, par [3], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance). As to claim 6, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the identifying whether the second AI model belongs to the optimal Pareto front comprises using a classifier (see page 4, par [3], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance). As to claim 7, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the at least two performance parameters are determined based on a plurality of objectives (see page 4, par [3], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance). As to claim 8, the combination Ma and Chen discloses the invention as claimed. In addition, Ma discloses the claimed wherein the classifier is a binary classifier (see abstract, propose a classification wise Pareto evolution approach for one-shot NAS, where an online classifier is trained to predict the dominance relationship between the candidate and constructed reference architectures, instead of using surrogates to fit the objective functions). As to claim 9, the combination Ma and Chen discloses the invention as claimed. In addition, Chen discloses the claimed wherein the classifier identifies the optimal Pareto front by identifying each probability that the second AI models belongs to the plurality of Pareto front (see page 4, par [3], the server according to the performance information of the third neural network and feedback function generating feedback value, comprising: the server compares the performance of the third neural network with the performance of at least one second neural network in the current Pareto front edge; the performance of the third neural network is better than the P second neural network in the current Pareto front edge, the server generates feedback value according to P and feedback function, wherein P is a positive integer, the feedback value is a positive number, P is larger, the feedback value is larger; under the condition that the performance of the third neural network is inferior to the Q second neural network in the current Pareto front edge, the server generates the feedback value according to Q and feedback function, wherein Q is a positive integer, the feedback value is negative number, Q is larger, the feedback value is smaller. In the implementation manner, when the third neural network sampled by the first circulating neural network is worse than the second neural network structure on the current pareto front edge, then to a negative feedback value, the first circulating neural network to sample the next new third neural network far away from the current third neural network. so as to guide the server to evolving to a pareto front edge with excellent performance from a Pareto front edge with poor performance). As to claims 10-18, claims 10-18 are electronic device for performing the method of claims 1-9 above. They are rejected under the same rationale. As to claims 19-20, claims 19-20 are non-transitory computer readable medium for executing the method of claims 1-9 above. They are rejected under the same rationale. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to JEAN M CORRIELUS whose telephone number is (571)272-4032. The examiner can normally be reached Monday-Friday 6:30a-10p(Midflex). 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 J 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. /JEAN M CORRIELUS/Primary Examiner, Art Unit 2159 June 15, 2026 Application/Control Number: 18/415,261 Page 2 Art Unit: 2159 Application/Control Number: 18/415,261 Page 3 Art Unit: 2159 Application/Control Number: 18/415,261 Page 4 Art Unit: 2159 Application/Control Number: 18/415,261 Page 5 Art Unit: 2159
Prosecution Insights