CTNF 18/292,244 CTNF 102081 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. Priority The present application has a provisional application No. 63/227,464 filed on July 30, 2021. The present application has a preliminary amendment filed on Jan, 25, 2024 with Claims 1-8, 18-19, 22-31 in 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, 4-8, 18-19, 24-28, 30-31 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Regarding Claim 1 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 1 is a method claim thus it falls into one of the four categories of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding independent claim 1, following limitations recite a judicial exception: “ generating, with at least one processor , a respective noise vector for a respective sample of the plurality of samples, the respective noise vector generated based on a length of the respective sample and a radius hyperparameter ” [Mathematical Calculations] – generating a vector respect to a number requires using mathematical calculations such as multivariable calculus and partial differentiation, which recites to an abstract idea. “ repeating, with at least one processor , for a target number of steps: adjusting, with at least one processor , the respective noise vector based on a step size hyperparameter; and projecting, with at least one processor , the respective noise vector to be within a boundary based on the radius hyperparameter if the respective noise vector was adjusted beyond the boundary after adjusting the respective noise vector ” [Mathematical Calculations] – repeating the adjustment and projection of the vector requires pure mathematical computation which recites to an abstract idea. “ adjusting, with at least one processor , the set of parameters of the deep learning model based on a gradient of a loss based on the respective noise vector ” [Mathematical Calculations] – adjusting the parameters based a gradient loss purely relies on mathematical computations such as a gradient descent and a backward propagation, which recites to an abstract idea. “ repeating, with at least one processor , the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples ” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 1, the claim recites additional elements of “ receiving, with at least one processor, a deep learning model comprising a set of parameters and a dataset comprising a plurality of samples ” Receiving data such as the model, its parameters and the dataset is merely data gathering recited at a high level of generality, thus is insignificant extra-solution activity (See MPEP 2106.05(g)). “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered an insignificant extra solution activity and at best the equivalent of a mere data gathering recited at a high level of generality and amount to receiving or transmitting data over a network, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains an insignificant extra solution activity. Also, the additional element [2] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. These limitations remain as they are even upon reconsideration. Even when considered in combination, the additional elements represent a mere instruction to apply an exception and an insignificant extra-solution activity, which cannot provide an inventive concept. Regarding Claim 4 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 4 is a dependent claim of 1, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 4, following limitations recite a judicial exception: “ generating the respective noise vector comprises generating the respective noise vector based on the following equation: PNG media_image1.png 77 161 media_image1.png Greyscale wherein δ comprises the noise vector, Li comprises the length of the respective sample, ϵ comprises the radius hyperparameter, and U(−ϵ, ϵ) comprises a uniform distribution from −ϵ to ϵ.” [Mathematical Formula] – generating a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 4 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 5 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 5 is a dependent claim of 1, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 5, following limitations recite a judicial exception: “ adjusting the respective noise vector comprises adjusting the respective noise vector based on the following equation: PNG media_image2.png 64 228 media_image2.png Greyscale wherein δ comprises the noise vector, α comprises the step size hyperparameter, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, ∇ δ is the gradient of δ, and yi comprises an expected output of the deep learning model” [Mathematical Formula] – adjusting a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 5 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 6 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 6 is a dependent claim of 1, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 6, following limitations recite a judicial exception: “ projecting the respective noise vector comprises projecting the respective noise vector based on the following equation: PNG media_image3.png 77 135 media_image3.png Greyscale wherein δ comprises the noise vector and ϵ comprises the radius hyperparameter” [Mathematical Formula] – projecting a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 6 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 7 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 7 is a dependent claim of 1, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 7, following limitations recite a judicial exception: “ adjusting the set of parameters comprises adjusting the set of parameters based on the following equation: PNG media_image4.png 41 195 media_image4.png Greyscale wherein δ comprises the noise vector, ϵ comprises the set of parameters, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, and yi comprises an expected output of the deep learning model.” [Mathematical Formula] – adjusting parameters using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 7 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 8 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 8 is a dependent claim of 1, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 8, following limitations recite a judicial exception: “ repeating, with at least one processor , for a target number of epochs, the repetition of the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples.” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 8, the claim recites additional elements of “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. This limitation remains a mere instruction to apply an exception even upon reconsideration. Even when considered in combination, the additional element represents a mere instruction to apply an exception, which cannot provide an inventive concept. Regarding Claim 18 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 18 is a system claim thus it falls into one of the four categories of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding independent claim 18, following limitations recite a judicial exception: “ generate a respective noise vector for a respective sample of the plurality of samples, the respective noise vector generated based on a length of the respective sample and a radius hyperparameter ” [Mathematical Calculations] – generating a vector respect to a number requires using mathematical calculations such as multivariable calculus and partial differentiation, which recites to an abstract idea. “ repeat for a target number of steps: adjust the respective noise vector based on a step size hyperparameter; and project the respective noise vector to be within a boundary based on the radius hyperparameter if the respective noise vector was adjusted beyond the boundary after adjusting the respective noise vector ” [Mathematical Calculations] – repeating the adjustment and projection of the vector requires pure mathematical computation which recites to an abstract idea. “ adjust the set of parameters of the deep learning model based on a gradient of a loss based on the respective noise vector ” [Mathematical Calculations] – adjusting the parameters based a gradient loss purely relies on mathematical computations such as a gradient descent and a backward propagation, which recites to an abstract idea. “ repeat the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples ” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 18, the claim recites additional elements of “ receive a deep learning model comprising a set of parameters and a dataset comprising a plurality of samples ” Receiving data such as the model, its parameters and the dataset is merely data gathering recited at a high level of generality, thus is insignificant extra-solution activity (See MPEP 2106.05(g)). “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered an insignificant extra solution activity and at best the equivalent of a mere data gathering recited at a high level of generality and amount to receiving or transmitting data over a network, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains an insignificant extra solution activity. Also, the additional element [2] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. These limitations remain as they are even upon reconsideration. Even when considered in combination, the additional elements represent a mere instruction to apply an exception and an insignificant extra-solution activity, which cannot provide an inventive concept. Regarding Claim 19 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 19 is a non-transitory medium claim thus it falls into one of the four categories of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding independent claim 19, following limitations recite a judicial exception: “ generate a respective noise vector for a respective sample of the plurality of samples, the respective noise vector generated based on a length of the respective sample and a radius hyperparameter ” [Mathematical Calculations] – generating a vector respect to a number requires using mathematical calculations such as multivariable calculus and partial differentiation, which recites to an abstract idea. “ repeat for a target number of steps: adjust the respective noise vector based on a step size hyperparameter; and project the respective noise vector to be within a boundary based on the radius hyperparameter if the respective noise vector was adjusted beyond the boundary after adjusting the respective noise vector ” [Mathematical Calculations] – repeating the adjustment and projection of the vector requires pure mathematical computation which recites to an abstract idea. “ adjust the set of parameters of the deep learning model based on a gradient of a loss based on the respective noise vector ” [Mathematical Calculations] – adjusting the parameters based a gradient loss purely relies on mathematical computations such as a gradient descent and a backward propagation, which recites to an abstract idea. “ repeat the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples ” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 19, the claim recites additional elements of “ receive a deep learning model comprising a set of parameters and a dataset comprising a plurality of samples ” Receiving data such as the model, its parameters and the dataset is merely data gathering recited at a high level of generality, thus is insignificant extra-solution activity (See MPEP 2106.05(g)). “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered an insignificant extra solution activity and at best the equivalent of a mere data gathering recited at a high level of generality and amount to receiving or transmitting data over a network, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains an insignificant extra solution activity. Also, the additional element [2] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. These limitations remain as they are even upon reconsideration. Even when considered in combination, the additional elements represent a mere instruction to apply an exception and an insignificant extra-solution activity, which cannot provide an inventive concept. Regarding Claim 24 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 24 is a dependent claim of 18, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 24, following limitations recite a judicial exception: “ generating the respective noise vector comprises generating the respective noise vector based on the following equation: PNG media_image1.png 77 161 media_image1.png Greyscale wherein δ comprises the noise vector, Li comprises the length of the respective sample, ϵ comprises the radius hyperparameter, and U(−ϵ, ϵ) comprises a uniform distribution from −ϵ to ϵ.” [Mathematical Formula] – generating a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 24 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 25 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 25 is a dependent claim of 18, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 25, following limitations recite a judicial exception: “ adjusting the respective noise vector comprises adjusting the respective noise vector based on the following equation: PNG media_image2.png 64 228 media_image2.png Greyscale wherein δ comprises the noise vector, α comprises the step size hyperparameter, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, ∇ δ is the gradient of δ, and yi comprises an expected output of the deep learning model” [Mathematical Formula] – adjusting a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 25 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 26 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 26 is a dependent claim of 18, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 26, following limitations recite a judicial exception: “ projecting the respective noise vector comprises projecting the respective noise vector based on the following equation: PNG media_image3.png 77 135 media_image3.png Greyscale wherein δ comprises the noise vector and ϵ comprises the radius hyperparameter” [Mathematical Formula] – projecting a vector using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 26 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 27 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 27 is a dependent claim of 18, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 27, following limitations recite a judicial exception: “ adjusting the set of parameters comprises adjusting the set of parameters based on the following equation: PNG media_image4.png 41 195 media_image4.png Greyscale wherein δ comprises the noise vector, ϵ comprises the set of parameters, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, and yi comprises an expected output of the deep learning model.” [Mathematical Formula] – adjusting parameters using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 27 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 28 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 28 is a dependent claim of 18, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 28, following limitations recite a judicial exception: “ repeat, for a target number of epochs, the repetition of the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples.” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 28, the claim recites additional elements of “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. This limitation remains a mere instruction to apply an exception even upon reconsideration. Even when considered in combination, the additional element represents a mere instruction to apply an exception, which cannot provide an inventive concept. Regarding Claim 30 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 30 is a dependent claim of 19, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 30, following limitations recite a judicial exception: “ generating the respective noise vector comprises generating the respective noise vector based on the following equation: PNG media_image1.png 77 161 media_image1.png Greyscale wherein δ comprises the noise vector, Li comprises the length of the respective sample, ϵ comprises the radius hyperparameter, and U(−ϵ, ϵ) comprises a uniform distribution from −ϵ to ϵ.” [Mathematical Formula] – generating a vector using the given formula recites to an abstract idea. “ adjusting the respective noise vector comprises adjusting the respective noise vector based on the following equation: PNG media_image2.png 64 228 media_image2.png Greyscale wherein δ comprises the noise vector, α comprises the step size hyperparameter, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, ∇ δ is the gradient of δ, and yi comprises an expected output of the deep learning model” [Mathematical Formula] – adjusting a vector using the given formula recites to an abstract idea. “ projecting the respective noise vector comprises projecting the respective noise vector based on the following equation: PNG media_image3.png 77 135 media_image3.png Greyscale wherein δ comprises the noise vector and ϵ comprises the radius hyperparameter” [Mathematical Formula] – projecting a vector using the given formula recites to an abstract idea. “ adjusting the set of parameters comprises adjusting the set of parameters based on the following equation: PNG media_image4.png 41 195 media_image4.png Greyscale wherein δ comprises the noise vector, ϵ comprises the set of parameters, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, and yi comprises an expected output of the deep learning model.” [Mathematical Formula] – adjusting parameters using the given formula recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? The claim 30 does not recite any additional elements other than abstract ideas, so it does not integrate into a practical application. Thus, this claim is directed to the abstract idea. Regarding Claim 31 Step 1 – whether the claim falls within any statutory category. See MPEP 2016.03 Claim 31 is a dependent claim of 19, thus it falls within the same category of statutory subject matter. Step 2A Prong 1 – whether the claim recites a judicial exception. See MPEP 2106.04, subsection II. Regarding dependent claim 31, following limitations recite a judicial exception: “ repeat, for a target number of epochs, the repetition of the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples.” [Mathematical Calculations] – the generating, the repeating, and the adjusting of the set of parameters purely relies on mathematical computations of mentioned above which recites to an abstract idea. Step 2A Prong 2 – whether the claim recites additional elements that integrate the exception into a practical application of the exception? Regarding Claim 28, the claim recites additional elements of “ at least one processor ” The processor is recited at a high level of generality and is merely adding words “apply it” to the judicial exception. (see MPEP 2106.05(f)) [Even when viewed in combination, the additional elements do no more than automate the mathematical calculations, using computer components as a tool, thus the claim as a whole does not integrate into a practical application.] Step 2B – whether the claim as a whole amount to significantly more than the judicial exception? I.e. Are there any additional elements (features/limitations/step) recited in the claim beyond the abstract idea? The claim does not provide an inventive concept (significantly more than the abstract idea). The claim is ineligible. As explained above, the additional element [1] is considered merely computer components that are just to store and execute code-based instructions which are considered a mere instruction to apply an exception and amount to storing and receiving information in memory, which is well-understood, routine, conventional activity (See MPEP 2106.05(d), subsection II). This limitation remains a mere instruction to apply an exception. This limitation remains a mere instruction to apply an exception even upon reconsideration. Even when considered in combination, the additional element represents a mere instruction to apply an exception, which cannot provide an inventive concept. 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. 07-23-aia AIA The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. 07-21-aia AIA Claim s 1-8, 18-19, 22-31 are rejected under 35 U.S.C. 103 as being unpatentable over Jiang et al. ( Jiang ), Non-Patent Literature listed in IDS filed on 03/13/2024, Version 4 of “SMART: Robust and Efficient Fine-Tuning for Pre-Trained Natural Language Models through Principled Regularized Optimization”, published on Feb 15, 2021, Pages: 21, in view of Yasunaga et al. ( Yasunaga ), Non-Patent Literature listed in IDS filed on 03/13/2024, , “Robust Multilingual Part-of-Speech Tagging via Adversarial Training”, published on Apr 20, 2018, Pages: 12 . As to independent Claim 1, Jiang teaches a computer-implemented method, comprising: receiving, with at least one processor, a deep learning model comprising a set of parameters and a dataset comprising a plurality of samples ( Jiang, Pg4, Section3.1, Paragraph1, Lines2, "Specifically, given the model PNG media_image5.png 27 60 media_image5.png Greyscale and n data points of the target task..., Jiang, Pg5, Section3.2, Paragraph1, Lines3, "Specifically, we use a pre-trained model as the initialization denoted by PNG media_image6.png 30 65 media_image6.png Greyscale ", Jiang, Pg7, Algorithm1, PNG media_image7.png 79 642 media_image7.png Greyscale , wherein the fine-tuning process of Jiang uses a pre-trained model with n data points or the corresponding dataset with the parameters shown in the input section that runs on a computer inherently indicates the data has been received with at least one processor, which is functionally equivalent to the claimed invention); repeating, with at least one processor, for a target number of steps: adjusting, with at least one processor, the respective noise vector based on a step size hyperparameter; and projecting, with at least one processor, the respective noise vector to be within a boundary based on the radius hyperparameter if the respective noise vector was adjusted beyond the boundary after adjusting the respective noise vector ( Jiang, Pg7, Algorithm1, Steps7-10 , PNG media_image8.png 800 966 media_image8.png Greyscale , wherein the algorithm shows the inner loop(step7-10) repeats Tx~ times to update the xi~, which is the noise or perturbed input vector which is functionally identical when it goes through the inner loop. Also, it updates or adjusts the corresponding noise vector using n, the learning rate parameter or the corresponding step size hyperparameter and the gradient gi~ in steps8-9, which is functionally identical to the corresponding adjustment. Also, the inequality in step9 represents that the projection of the noise vector once it goes beyond the boundary of e-ball such that it adjusts the vector to fit back into the boundary of e. Thus, it is functionally equivalent to the claimed invention); adjusting, with at least one processor, the set of parameters of the deep learning model based on a gradient of a loss based on the respective noise vector (Pg4, Section 3.1, Paragraph1, Lines4-10, "Our method essentially solves the following optimization for fine-tuning: PNG media_image9.png 34 629 media_image9.png Greyscale where L(theta) is the loss function … and Rs(theta) is the smoothness-inducing adversarial regularizer. Here we define Rs(theta) as PNG media_image10.png 78 426 media_image10.png Greyscale ”, Pg6, Line3, "we need to solve it using SGD-type algorithms such as ADAM", Jiang, Pg7, Algorithm1, Step11 , wherein the optimization process using the F(theta) or equation 1 involves both L(theta) and Rs(theta) which are the corresponding loss based on the noise vector, xi~, to adjust the parameters while the ADAM is the SGD-type or stochastic gradient descent which adjusts the parameters shown in the step11 inherently updating the parameters based on the gradient of a loss , thus it is functionally equivalent to the claimed invention); and repeating, with at least one processor, the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples (Jiang, Pg7, Algorithm1, PNG media_image7.png 79 642 media_image7.png Greyscale , wherein T shows that the entire algorithm iterates T times to do the corresponding repetition, generations, adjustments of the data set X or the corresponding set of parameters for each sample of the plurality of samples, which is functionally equivalent to the claimed invention.) Jiang , however, does not teach generating, with at least one processor, a respective noise vector for a respective sample of the plurality of samples, the respective noise vector generated based on a length of the respective sample and a radius hyperparameter. From the same field of endeavor, Yasunaga teaches this limitation ( Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Lines9-12, "To prepare an adversarial example, we aim to generate the worst-case perturbation of a small bounded norm e that maximizes the loss function L of the current model:" Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Paragraph3, Lines1-13, "While Miyato et al. (2017) set the norm of a perturbation e (Eq 2) to be a fixed value for all input sentences, to generate adversarial examples for an entire sentence of a variable length and to include character embeddings besides word embeddings, we make the perturbation size adaptive to the dimension of the concatenated input embedding s in R^D. We set e to be a(sqrt(D)) (i.e., proportional to sqrt(D)), as the expected squared norm of s after the embedding normalization is D. The scaling factor a is selected from {0.001, 0.005, 0.01, 0.05, 0.1} based on the development performance in each treebank" , wherein generating the noise vectors which are bounded by e such that e is based on the length of the sentences instead of a fixed value so e is dynamically adapted. Also, the scaling factor a or the corresponding radius hyperparameter is used to scale the noise vector accordingly, which is functionally equivalent to the claimed invention. ) Jiang and Yasunaga are analogous to the claimed invention as they are from the same field of endeavor of regularizing deep learning models for natural language processing via adversarial training to improve model robustness and generalization performance against input perturbations during a fine-tuning stage. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to combine the iterative noise adjustment within the trust-region boundary of Jiang with the dynamic scaling the allowable radius boundary of the noise vector based on the measured variable length of the input text sample of Yasunaga . The motivation is as recited by Yasunaga ( Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Paragraph3, Lines1-8, "While Miyato et al. (2017) set the norm of a perturbation e (Eq 2) to be a fixed value for all input sentences, to generate adversarial examples for an entire sentence of a variable length and to include character embeddings besides word embeddings, we make the perturbation size adaptive to the dimension of the concatenated input embedding s in R^D") such that applying uniform, static perturbation radius across text sequences of varying lengths in NLP tasks fails to account for the inherent variance in sentence length, which inherently distorts the regularization density and the actual intensity of the adversarial attack across different inputs. As to dependent Claim 2, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. Jiang further teaches the method of claim 1, wherein the deep learning model comprises a natural language processing (NLP) model ( Jiang, Pg14, Section6 Conclusion, Lines5-7, "Our empirical results suggest that SMART improves the performance on many NLP benchmarks (e.g., GLUE, SNLI, SciTail, ANLI) with the state-of-the-art pre-trained models (e.g.,BERT, MT-DNN, RoBERTa).") As to dependent Claim 3, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1 and 2. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance of the deep learning model such as a natural language processing model. Jiang further teaches the method of claim 2, wherein the NLP model comprises a Bidirectional Encoder Representations from Transformers (BERT) model ( Jiang, Pg14, Section6 Conclusion, Lines5-7, "Our empirical results suggest that SMART improves the performance on many NLP benchmarks (e.g., GLUE, SNLI, SciTail, ANLI) with the state-of-the-art pre-trained models (e.g.,BERT, MT-DNN, RoBERTa).") As to dependent Claim 4, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. While Jiang does not teach the following limitation, but from the same field of endeavor, Yasunaga teaches the method of claim 1, wherein generating the respective noise vector comprises generating the respective noise vector based on the following equation: PNG media_image1.png 77 161 media_image1.png Greyscale wherein δ comprises the noise vector, Li comprises the length of the respective sample, ϵ comprises the radius hyperparameter, and U(−ϵ, ϵ) comprises a uniform distribution from −ϵ to ϵ ( Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Lines9-12, "To prepare an adversarial example, we aim to generate the worst-case perturbation of a small bounded norm e that maximizes the loss function L of the current model:", Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Paragraph3, Lines1-13, "While Miyato et al. (2017) set the norm of a perturbation e (Eq 2) to be a fixed value for all input sentences, to generate adversarial examples for an entire sentence of a variable length and to include character embeddings besides word embeddings, we make the perturbation size adaptive to the dimension of the concatenated input embedding s in R^D. We set e to be a(sqrt(D)) (i.e., proportional to sqrt(D)), as the expected squared norm of s after the embedding normalization is D. The scaling factor a is selected from {0.001, 0.005, 0.01, 0.05, 0.1} based on the development performance in each treebank" , wherein as mentioned in Claim1, Yasunaga shows generating adversarial vectors which are bounded by e which depends on the variable length of an entire sentence with the scalling factor a(the corresponding radius hyperparameter) such that it is functionally equivalent to the claimed invention of using (-e,e) as the boundary to create the noise vector respect to the length of the sentence. ) Jiang and Yasunaga are analogous to the claimed invention as they are from the same field of endeavor of regularizing deep learning models for natural language processing via adversarial training to improve model robustness and generalization performance against input perturbations during a fine-tuning stage. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to combine the iterative noise adjustment within the trust-region boundary of Jiang with the dynamic scaling the allowable radius boundary of the noise vector based on the measured variable length of the input text sample of Yasunaga . The motivation is as recited by Yasunaga ( Yasunaga, Pg4, Left Column, Section: Generating adversarial examples, Paragraph3, Lines1-8, "While Miyato et al. (2017) set the norm of a perturbation e (Eq 2) to be a fixed value for all input sentences, to generate adversarial examples for an entire sentence of a variable length and to include character embeddings besides word embeddings, we make the perturbation size adaptive to the dimension of the concatenated input embedding s in R^D") such that applying uniform, static perturbation radius across text sequences of varying lengths in NLP tasks fails to account for the inherent variance in sentence length, which inherently distorts the regularization density and the actual intensity of the adversarial attack across different inputs. As to dependent Claim 5, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. Jiang further teaches the method of claim 1, wherein adjusting the respective noise vector comprises adjusting the respective noise vector based on the following equation: PNG media_image2.png 64 228 media_image2.png Greyscale wherein δ comprises the noise vector, α comprises the step size hyperparameter, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, ∇ δ is the gradient of δ, and yi comprises an expected output of the deep learning model ( Jiang, Pg7, Algorithm1, Steps7-10 PNG media_image8.png 800 966 media_image8.png Greyscale , wherein as mentioned in Claim 1, the inner loop indicates that the noise vector, xi~, is adjusted using n, the learning rate parameter or the corresponding step size hyperparameter and the gradient gi~ in step8. The step8 mathematically shows the gradient normalization such that the gradient gi of the current loss function is used to leave the most adversarial features. Also, the step size parameter or the learning rate factor n is multiplied in step9 to gradually make the noise vector more adversarial, which is functionally equivalent to the claimed invention.) As to dependent Claim 6, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. Jiang further teaches the method of claim 1, wherein projecting the respective noise vector comprises projecting the respective noise vector based on the following equation: PNG media_image3.png 77 135 media_image3.png Greyscale wherein δ comprises the noise vector and ϵ comprises the radius hyperparameter ( Jiang, Pg7, Algorithm1, Notation and Steps9, " PNG media_image11.png 31 41 media_image11.png Greyscale denotes the projection to A", Pg21, Section 8 Hyperparameters, third bullet point, "p = infinity makes the size of perturbation constraint to be the same regardless of the number of dimensions. For p=2, adversrial perturbation is sensitive to the number of dimensions(A higher dimension usually requires a larger perturbation), espeicially for sentences with different length. As a result, we need to make less tuning effort for p = infinity. For other values of p, the associated projections are computationally inefficient" , wherein the adjust input PNG media_image12.png 34 89 media_image12.png Greyscale of the inner loop goes beyond the specified trust-region boundary, the algorithm will do the projection of the Algorithm forcing the noise back into the designated boundary set. While Jiang's default algorithm utilizes the p = infinity norm boundary, Jiang explicitly acknowledges and analyzes the functional alternative of utilizing a spherical boundary (p=2 norm, which corresponds to the claimed invention) along with its associated projection mechanics, thus it is functionally equivalent to the claimed invention. ) As to dependent Claim 7, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. Jiang further teaches the method of claim 1, wherein adjusting the set of parameters comprises adjusting the set of parameters based on the following equation: PNG media_image4.png 41 195 media_image4.png Greyscale wherein δ comprises the noise vector, ϵ comprises the set of parameters, l( ) comprises a loss function, ƒθ( ) comprises an output of the deep learning model, and yi comprises an expected output of the deep learning model (Pg4, Section 3.1, Paragraph1, Lines4, "Our method essentially solves the following optimization for fine-tuning: PNG media_image9.png 34 629 media_image9.png Greyscale where L(theta) is the loss function … and Rs(theta) is the smoothness-inducing adversarial regularizer. Here we define Rs(theta) as PNG media_image10.png 78 426 media_image10.png Greyscale ", Pg6, Line3, "we need to solve it using SGD-type algorithms such as ADAM", Jiang, Pg7, Algorithm1, Step11 , wherein as mentioned in Claim1, Jiang discloses optimizing model parameters, theta, by minimizing an objective function F(theta) that incorporates a loss component evaluated under the worst-case adversarial perturbation (xi~), where Rs(theta) represents the regularization term computed based on the perturbed input xi~, found during the inner loop optimization. Jiang also discloses that he parameter optimization is driven by gradient-based stochastic gradient descent (SGD) variants, which inherently compute and apply the gradient of the loss with respect to the parameters. Also, as shown in the step 11 of the algorithm, Jiang shows explicit step-by-step algorithmic execution where the parameter set theta is iteratively adjusted using the SGD, or ADAM.) As to dependent Claim 8, The combination of Jiang and Yasunaga, as mentioned above, teaches all the limitations of Claim 1. The combination teaches about the fine-tuning process which creates a perturbed or an adversarial noise through repeated tuning procedures, then which is applied to the regular gradient descent for multiple epochs until it reaches the desired accuracy and performance. Jiang further teaches the method of claim 1, further comprising: repeating, with at least one processor, for a target number of epochs, the repetition of the generating, the repeating for the target number of steps, and the adjusting of the set of parameters for each sample of the plurality of samples ( Jiang, Pg7, Algorithm1, Input PNG media_image7.png 79 642 media_image7.png Greyscale , wherein as mentioned in Claim1, utilizing a "target number of epochs" as a termination or repetition criterion is functionally equivalent to the entire algorithm iterating T times to do the corresponding generations and adjustments before its termination.) As to independent Claim 18, it is a system claim that contains similar limitations of Claim 1 and thus rejected under the same rationale. As to independent Claim 19, it is a product claim that contains similar limitations of Claim 1 and thus rejected under the same rationale. As to dependent Claim 22, it is a system claim that contains similar limitations of Claim 2 and thus rejected under the same rationale. As to dependent Claim 23, it is a system claim that contains similar limitations of Claim 3 and thus rejected under the same rationale. As to dependent Claim 24, it is a system claim that contains similar limitations of Claim 4 and thus rejected under the same rationale. As to dependent Claim 25, it is a system claim that contains similar limitations of Claim 5 and thus rejected under the same rationale. As to dependent Claim 26, it is a system claim that contains similar limitations of Claim 6 and thus rejected under the same rationale. As to dependent Claim 27, it is a system claim that contains similar limitations of Claim 7 and thus rejected under the same rationale. As to dependent Claim 28, it is a system claim that contains similar limitations of Claim 8 and thus rejected under the same rationale. As to dependent Claim 29, it is a product claim that contains similar limitations of Claims 2 and 3 and thus rejected under the same rationale. As to dependent Claim 30, it is a product claim that contains similar limitations of Claims 4-7 and thus rejected under the same rationale. As to dependent Claim 31, it is a product claim that contains similar limitations of Claim 8 and thus rejected under the same rationale . Conclusion 07-96 AIA The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Dathathri et al., Non-Patent Literature, “PLUG AND PLAY LANGUAGE MODELS: A SIMPLE APPROACH TO CONTROLLED TEXT GENERATION”, published on 2020, 34 Pages Howard et al., Non-Patent Literature, “Universal Language Model Fine-Tuning for Text Classification”, published on 2018, 12 Pages Any inquiry concerning this communication or earlier communications from the examiner should be directed to DONG YOON JUNG whose telephone number is (571)270-0198. The examiner can normally be reached 8am-5pm. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Cesar Paula can be reached at (571) 272-4128. 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. /DONG YOON JUNG/Examiner, Art Unit 2145 /CESAR B PAULA/Supervisory Patent Examiner, Art Unit 2145 Application/Control Number: 18/292,244 Page 2 Art Unit: 2145 Application/Control Number: 18/292,244 Page 3 Art Unit: 2145 Application/Control Number: 18/292,244 Page 4 Art Unit: 2145 Application/Control Number: 18/292,244 Page 5 Art Unit: 2145 Application/Control Number: 18/292,244 Page 6 Art Unit: 2145 Application/Control Number: 18/292,244 Page 7 Art Unit: 2145 Application/Control Number: 18/292,244 Page 8 Art Unit: 2145 Application/Control Number: 18/292,244 Page 9 Art Unit: 2145 Application/Control Number: 18/292,244 Page 10 Art Unit: 2145 Application/Control Number: 18/292,244 Page 11 Art Unit: 2145 Application/Control Number: 18/292,244 Page 12 Art Unit: 2145 Application/Control Number: 18/292,244 Page 13 Art Unit: 2145 Application/Control Number: 18/292,244 Page 14 Art Unit: 2145 Application/Control Number: 18/292,244 Page 15 Art Unit: 2145 Application/Control Number: 18/292,244 Page 16 Art Unit: 2145 Application/Control Number: 18/292,244 Page 17 Art Unit: 2145 Application/Control Number: 18/292,244 Page 18 Art Unit: 2145 Application/Control Number: 18/292,244 Page 19 Art Unit: 2145 Application/Control Number: 18/292,244 Page 20 Art Unit: 2145 Application/Control Number: 18/292,244 Page 21 Art Unit: 2145 Application/Control Number: 18/292,244 Page 22 Art Unit: 2145 Application/Control Number: 18/292,244 Page 23 Art Unit: 2145 Application/Control Number: 18/292,244 Page 24 Art Unit: 2145 Application/Control Number: 18/292,244 Page 25 Art Unit: 2145 Application/Control Number: 18/292,244 Page 26 Art Unit: 2145 Application/Control Number: 18/292,244 Page 27 Art Unit: 2145 Application/Control Number: 18/292,244 Page 28 Art Unit: 2145 Application/Control Number: 18/292,244 Page 29 Art Unit: 2145 Application/Control Number: 18/292,244 Page 30 Art Unit: 2145 Application/Control Number: 18/292,244 Page 31 Art Unit: 2145 Application/Control Number: 18/292,244 Page 32 Art Unit: 2145 Application/Control Number: 18/292,244 Page 33 Art Unit: 2145 Application/Control Number: 18/292,244 Page 34 Art Unit: 2145 Application/Control Number: 18/292,244 Page 35 Art Unit: 2145 Application/Control Number: 18/292,244 Page 36 Art Unit: 2145 Application/Control Number: 18/292,244 Page 37 Art Unit: 2145 Application/Control Number: 18/292,244 Page 38 Art Unit: 2145
Prosecution Insights