DETAILED ACTION
This action is in response the communications filed on 02/25/2026 in which no claims are amended, and therefore claims 1-19 are pending.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 02/25/2026 has been entered.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
-
Claims 1-19 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more
Step 1: Claims 1-17 recite a method. Claim 18 recites a system comprising computers and storage devices. Therefore, claims 1-17 are directed to a process, and claim 18 is directed to a machine.
With respect to claims 1, 18 and 19:
2A Prong 1: The claim recites a judicial exception.
the method comprising repeatedly performing a second or higher order update operation at successive time steps, the second or higher order update operation comprising: based on current values for the first and second numerical parameters, generating at least one set of first intermediate values for the first numerical parameters using gradients of a first loss component, and at least one set of second intermediate values for the second numerical parameters using gradients of a second loss component; (mathematical concept - mathematical calculation; see also spec [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6)”)
generating respective second or higher order updates to the first and second numerical parameters based on gradients of the loss components with respect to the first and second numerical parameters both for the current values of the first and second numerical parameters and for the first and second intermediate values of the first and second numerical parameters; and (mathematical concept - mathematical calculation; see also spec , [0075]-[0076] “In step 203, updates (that is respective update amounts) are obtained to the current values (θk, ϕk) of the first and second numerical parameters. These updates are second or higher order. That is, they are a function both of the current values of the first and second numerical values and also of the set(s) of first and second intermediate data. Thus, each update encodes and benefits from higher derivatives of the loss function… This is used in step 203 to produce the two values called here… h/2 (v(θk, ϕk) + v(θ~k, ϕ~k)) [second or higher order updates]”)
updating the current values of the first and second numerical parameters by corresponding updates (mathematical concept - mathematical calculation; see also spec , [0077] “In step 204, the current values of the first and second numerical parameters are updated by adding the respective updates, to give: (θk+1 ϕk+1) = (θk ϕk) + h/2 (v(θk,ϕk) + v(θ~k,ϕ~k)).(7)”)
2A Prong 2: The judicial exception is not integrated into a practical application.
training an adversarial machine learning model based on a plurality of numerical parameters, the numerical parameters comprising one or more first numerical parameters and one or more second numerical parameters, the training being performed to minimize an objective function having a plurality of loss components, at least one of the loss components being a function of both the first and second numerical parameters, wherein optimizing one of the loss components with respect to the numerical parameters tends to move another of the loss components away from its optimal value (mere instructions to apply an exception – MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception; high level recitation of training a model with an objective function)
(claims 18 and 19) one or more computers, one or more storage devices, instructions (mere instructions to apply an exception - MPEP 2106.05(f), (2) Whether the claim invokes computer; generic computer components)
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is directed to an abstract idea.
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
training an adversarial model based on a plurality of numerical parameters, the numerical parameters comprising one or more first numerical parameters and one or more second numerical parameters, the training being performed to minimize an objective function having a plurality of loss components, at least one of the loss components being a function of both the first and second numerical parameters, wherein optimizing one of the loss components with respect to the numerical parameters tends to move another of the loss components away from its optimal value (mere instructions to apply an exception – MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception; high level recitation of training a model with an objective function)
(claims 18 and 19) one or more computers, one or more storage devices, instructions(mere instructions to apply an exception - MPEP 2106.05(f), (2) Whether the claim invokes computer; generic computer components)
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
With respect to claim 2:
2A Prong 1: The claim recites a judicial exception.
in which said generating at least one set of first intermediate values for the first numerical parameters using gradients of the first loss component, and at least one set of second intermediate values for the second numerical parameters using gradients of the second loss component, is performed in one or more successive intermediate steps (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6)”)
in each intermediate step, a set of intermediate values for the numerical parameters is generated comprising a respective said set of first intermediate values for the first numerical parameters and a respective said set of second intermediate values for the second numerical parameters (mathematical concepts, in light of [0074] θ~k for θ and ϕ~k for ϕ)
With respect to claim 3:
2A Prong 1: The claim recites a judicial exception.
in which, in the first intermediate step: each first intermediate value is derived by adjusting the current value of the first numerical parameter by a respective amount indicative of the gradient of the first loss component with respect to the first numerical parameter for the current values of the numerical parameters (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) where the vector v (θk, φk) is given by ν (θk,ϕk) = -[α ∂lD/∂θ,β ∂lG/∂(ϕ)] [a respective amount indicative of the gradient]”)
each second intermediate value is derived by adjusting the current value of the second numerical parameter by a respective amount indicative of the gradient of the second loss component with respect to the second numerical parameter for the current values of the numerical parameters (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) where the vector v (θk, φk) is given by ν (θk,ϕk) = -[α ∂lD/∂θ,β ∂lG/∂(ϕ)] [a respective amount indicative of the gradient]”)
With respect to claim 4:
2A Prong 1: The claim recites a judicial exception.
in which there are a plurality of said intermediate steps, and each of the intermediate steps except a first comprises evaluating the gradient of the first loss component with respect to the first numerical parameters and the gradient of the second loss component with respect to the second numerical parameters, each of the evaluations being performed for the intermediate values generated in the preceding intermediate step (mathematical concepts, [0075] “In step 203, updates (that is respective update amounts) are obtained to the current values (θk, ϕk) of the first and second numerical parameters. These updates are second or higher order. That is, they are a function both of the current values of the first and second numerical values and also of the set(s) of first and second intermediate data. Thus, each update encodes and benefits from higher derivatives of the loss function. [the gradient of both the 1st and 2nd loss functions with respect to the 1st and 2nd numerical parameters]”)
With respect to claim 5:
2A Prong 1: The claim recites a judicial exception.
in which: an update for each first numerical parameter is a sum of a term indicative of the gradient of the first loss component with respect to the first numerical parameter for the current values of the numerical parameters, and, for each of the intermediate steps, a term indicative of the gradient the first loss component with respect to the first numerical parameter for the corresponding intermediate values of the numerical parameters (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) [a sum of a term indicative of the gradient] where the vector v (θk, φk) is given by ν (θk,ϕk) = -[α ∂lD/∂θ,β ∂lG/∂(ϕ)] [a term indicative of the gradient]”)
an update for each second numerical parameter is a sum of a term indicative of the gradient of the second loss component with respect to the second numerical parameter for the current values of the numerical parameters, and, for each of the intermediate steps, a term indicative of the gradient of the second loss component with respect to the second numerical parameter for the corresponding intermediate values of the numerical parameters (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) [a sum of a term indicative of the gradient] where the vector v (θk, φk) is given by ν (θk,ϕk) = -[α ∂lD/∂θ,β ∂lG/∂(ϕ)] [a term indicative of the gradient]”)
With respect to claim 6:
2A Prong 1: The claim recites a judicial exception.
in which there is only one said intermediate step, and the update for each first numerical parameter is indicative of an average of (i) the gradient of the first loss component with respect to the first numerical parameter for the current values of the numerical parameters, and (ii) the gradient the first loss component with respect to the first numerical parameter for the set of intermediate values of the numerical parameters (mathematical concepts, [0016] “This may be expressed mathematically as: (θk+1 ϕk+1) = (θk ϕk) + h/2 (v(θk,ϕk) + v(θ~k, ϕ~k)) [average of (i) v(θk) and (ii) v(θ~k)] (3) where θ~k and ϕ~k are the first intermediate values given by Eqn. (1).”; [0076] “This is used in step 203 to produce the two values called here… h/2 (v(θk, ϕk) + v(θ~k, ϕ~k)) [average of (i) v(θk) and (ii) v(θ~k)]”)
the update for each second numerical parameter is indicative of an average of (i) the gradient of the second loss component with respect to the second numerical parameter for the current values of the numerical parameters, and (ii) the gradient the second loss component with respect to the second numerical parameter for the set of intermediate values of the numerical parameters (mathematical concepts, [0016] “This may be expressed mathematically as: (θk+1 ϕk+1) = (θk ϕk) + h/2 (v(θk,ϕk) + v(θ~k, ϕ~k)) [average of (i) v(ϕk) and (ii) v(ϕ~k)] (3) where θ~k and ϕ~k are the first intermediate values given by Eqn. (1).”; [0076] “This is used in step 203 to produce the two values called here… h/2 (v(θk, ϕk) + v(θ~k, ϕ~k)) [average of (i) v(ϕk) and (ii) v(ϕ~k)]”)
With respect to claim 7:
2A Prong 1: The claim recites a judicial exception.
in which there are a plurality of intermediate steps, and in each intermediate step but the first: each first intermediate value is derived by adjusting the current value of the first numerical parameter by a respective amount indicative of the gradient of the first loss component with respect to the first numerical parameter for the set of intermediate values of the numerical parameters derived in the preceding intermediate step (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) [adjusting the current value of the first numerical parameter θk] where the vector v (θk, φk) is given by ν (θk, ϕk) = -[α ∂lD/∂θ,β ∂lG/∂(ϕ)] [by a respective amount indicative of the gradient] evaluated at (θk, ϕk).”)
each second intermediate value is derived by adjusting the current value of the second numerical parameter by a respective amount indicative of the gradient of the second loss component with respect to the second numerical parameter for the set of intermediate values of the numerical parameters derived in the preceding intermediate step (mathematical concepts, [0074] “the first intermediate values θ~k for the first parameters, and second intermediate values ϕ~k for the second numerical parameters may be given by: (θ˜k ϕ˜k) = (θk ϕk) + hν(θk, ϕk) (6) [adjusting the current value of the second numerical parameter ϕk] where the vector v (θk, φk) is given by ν (θk, ϕk) = -[α ∂lD/∂θ, β ∂lG/∂(ϕ)] [by a respective amount indicative of the gradient] evaluated at (θk, ϕk).”)
With respect to claim 8:
2A Prong 1: The claim recites a judicial exception.
in which there are three intermediate steps, and the update for each first numerical parameter is indicative of an average of (i) the gradient of the first loss component with respect to the first numerical parameter for the current values of the numerical parameters, (ii) twice the gradient the first loss component with respect to the first numerical parameter for the first set of intermediate values of the numerical parameters, (iii) twice the gradient the first loss component with respect to the first numerical parameter for the second set of intermediate values of the numerical parameters; and (iv) the gradient the first loss component with respect to the first numerical parameter for a third set of intermediate values of the numerical parameters (mathematical concepts, [0081] “In sub-step 301 of step 202, first intermediate values are defined. This may be done by defining: v1 = v(θk, φk) [(i) the gradient of lD wrt. θ for θk]”; [0082] “ν2 = ν(θk + h/2(ν1)θ, ϕk + h/2(ν1)ϕ) [(ii) the gradient of lD wrt. θ for v1]”; [0084] “ν3 = ν(θk + h/2(ν2)θ, ϕk + h/2(ν2)ϕ) [(iii) the gradient of lD wrt. θ for v2]”; [0086]-[0087] “intermediate values is generated… (θk + h(v3)θ, ϕk + h(v3)ϕ) [for a third set of intermediate values]… v4 = v(θk +h(v3)θ, ϕk + h(v3)ϕ) [(iv) the gradient of lD wrt. θ for v3]”; [0089] “h/6 (ν1 + 2ν2 + 2ν3 + ν4).” [an average of (i) + 2*(ii) + 2*(iii) + (iv)])
the update for each second numerical parameter is indicative of an average of (i) the gradient of the second loss component with respect to the second numerical parameter for the current values of the numerical parameters, (ii) twice the gradient the second loss component with respect to the second numerical parameter for the first set of intermediate values of the numerical parameters, (iii) twice the gradient the second loss component with respect to the second numerical parameter for the second set of intermediate values of the numerical parameters; and (iv) the gradient the second loss component with respect to the second numerical parameter for a third set of intermediate values of the numerical parameters (mathematical concepts, [0081] “In sub-step 301 of step 202, first intermediate values are defined. This may be done by defining: v1 = v(θk, φk) [(i) the gradient of lD wrt. φ for φk]”; [0082] “ν2 = ν(θk + h/2(ν1)θ, ϕk + h/2(ν1)ϕ) [(ii) the gradient of lD wrt. φ for v1]”; [0084] “ν3 = ν(θk + h/2(ν2)θ, ϕk + h/2(ν2)ϕ) [(iii) the gradient of lD wrt. φ for v2]”; [0086]-[0087] “intermediate values is generated… (θk + h(v3)θ, ϕk + h(v3)ϕ) [for a third set of intermediate values]… v4 = v(θk +h(v3)θ, ϕk + h(v3)ϕ) [(iv) the gradient of lD wrt. φ for v3]”; [0089] “h/6 (ν1 + 2ν2 + 2ν3 + ν4).” [an average of (i) + 2*(ii) + 2*(iii) + (iv)])
With respect to claim 9:
2A Prong 1: The claim recites a judicial exception.
in which the update process further comprises a regularization update, to at least one of the first numerical parameters and the second numerical parameters, the regularization update being performed by subtracting a regularization amount from the corresponding one of the updated first numerical parameters and the updated second numerical parameters (mathematical concepts, [0078] “In optional step 205, a regularization amount is subtracted from the new current values of the first and/or second numerical parameters… by removing the amount hλgθ where λ is a regularization multiplier, and gθ = ∇Θ ||∂lG/∂ϕ❘θk,ϕk||2”)
With respect to claim 10:
2A Prong 1: The claim recites a judicial exception.
in which the regularization amount is a positive number, the number being indicative in a case of the first numerical parameters of magnitude of gradient of the first loss component with respect to the first numerical parameters for the updated numerical parameters, and in a case of the second numerical parameters of magnitude of the gradient of the second loss component with respect to the second numerical parameters for the updated numerical parameters (mathematical concepts, [0078] “In optional step 205, a regularization amount is subtracted from the new current values of the first and/or second numerical parameters… by removing the amount hλgθ where λ is a regularization multiplier, and gθ = ∇Θ ||∂lG/∂ϕ❘θk,ϕk||2”; 2-norm is the positive square root of the sum of the squared absolute values of the components of a vector)
With respect to claim 11:
2A Prong 1: The claim recites a judicial exception.
in which the regularization amount, in a case of the first numerical parameters is proportional to a square of the gradient of the first loss component with respect to the first numerical parameters for the current numerical parameters, and in a case of the second numerical parameters is proportional to a square the gradient of the second loss component with respect to the second numerical parameters for the current numerical parameters (mathematical concepts, [0078] “In optional step 205, a regularization amount is subtracted from the new current values of the first and/or second numerical parameters… by removing the amount hλgθ where λ is a regularization multiplier, and gθ = ∇Θ ||∂lG/∂ϕ❘θk,ϕk||2”; 2-norm is the positive square root of the sum of the squared absolute values of the components of a vector)
With respect to claim 12:
2A Prong 2: The judicial exception is not integrated into a practical application.
in which an adaptive system comprises a generative neural network configured to generate samples based on one or more latent values, and a discriminator neural network configured to distinguish between samples generated by the generative neural network and samples from a sample distribution which are not generated by the generative neural network (MPEP 2106.05(f), more instructions to apply an exception, (3) The particularity or generality of the application of the judicial exception; In a Generative Adversarial Network (GAN), the generator creates synthetic data from real data. The discriminator receives both real data and the generator's fake data as input and determining the probability of each sample being real.)
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is directed to an abstract idea.
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
in which an adaptive system comprises a generative neural network configured to generate samples based on one or more latent values, and a discriminator neural network configured to distinguish between samples generated by the generative neural network and samples from a sample distribution which are not generated by the generative neural network (MPEP 2106.05(f) more instructions to apply an exception, (3) The particularity or generality of the application of the judicial exception)
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
With respect to claim 13:
2A Prong 1: The claim recites a judicial exception.
in which the discriminator neural network is defined by the first numerical parameters, and the generative neural network defined by the second numerical parameters (mathematical concepts, [0025] “The first numerical parameters θ may be the parameters of the discriminator network, and the second numerical parameters ϕ may be the parameters of the generative neural network”; [0067] “G(z; ϕ) is a generated sample 112… generated by the generative neural network… D(G(z; ϕ); θ) is the likelihood assigned by the discriminator neural network system”)
With respect to claim 14:
2A Prong 1: The claim recites a judicial exception.
in which both the loss components are indicative of a sum of (i) an average over the distribution of the latent values of a term indicative of the output of the discriminator network upon receiving as an input the output of the generative neural network generated based on the latent values, and (ii) an average over a training set of data items of a term indicative of the output of the discriminator network upon receiving as an input a data item from the training set (mathematical concepts, [0069] “Furthermore, the loss function may be an average over the sample distribution p(x) and the latent distribution p(z), where these averages are typically evaluated by considering many possible realizations of real samples x and latent values z. Thus, the loss function may be: J(θ,ϕ) = Ex˜p(x)[log(D(x;θ))] + Ez˜p(z)[log(1−D(G(z;ϕ));θ))] (4) [a sum of (i) Ez˜p(z)[log(1−D(G(z;ϕ));θ))] an average over p(z) of D(G(z;ϕ)) and (ii) Ex˜p(x)[log(D(x;θ))]] an average over p(x) of D(x;θ)”)
With respect to claim 15:
2A Prong 1: The claim recites a judicial exception.
which the minimizing the first loss component with respect to the first parameters corresponds to (mathematical concepts, in light of specification [0070] “In a general form which covers all these possible loss functions, the overall loss function (objective function) of the discriminator-generator pair can be denoted by: l(θ,ϕ)=[lD(θ,ϕ), lD(θ,ϕ)] (5) [lD(θ,ϕ) the first loss component] The training process amounts to minimizing the loss component lD(θ, ϕ) with respect to θ [1st parameter], and the loss component lG(θ, ϕ) with respect to ϕ”)
maximizing a measure of a difference between (i) an average over the distribution of the latent values of a term indicative of the output of the discriminator network upon receiving as an input the output of the generative neural network generated based on the latent values, and (ii) an average over a training set of data items of a term indicative of the output of the discriminator network upon receiving as an input a data item from the training set (mathematical concepts, [0069] “ Training is thus finding the (θ, ϕ) which is maxθ minϕ J(θ, ϕ). [maxθ: maximizing a measure] In some cases, the problem is transformed into one in which the objective function is asymmetric, e.g. J(θ,ϕ) = Ex˜p(x)[log(D(x;θ))] + Ez˜p(z)[−log(D(G(z;ϕ));θ))] [a difference between (i) Ez˜p(z)[log(D(G(z;ϕ));θ))] an average over p(z) of D(G(z;ϕ)) and (ii) Ex˜p(x)[log(D(x;θ))]] an average over p(x) of D(x;θ))
minimizing the second loss component with respect to the second parameters corresponds to minimizing said measure of the difference (mathematical concepts, [0069] “Training is thus finding the (θ, ϕ) which is maxθ minϕ J(θ, ϕ). [minϕ: minimizing the second loss component] In some cases, the problem is transformed into one in which the objective function is asymmetric, e.g. J(θ,ϕ) = Ex˜p(x)[log(D(x;θ))] + Ez˜p(z)[−log(D(G(z;ϕ));θ))] [said measure of the difference] )
With respect to claim 16:
2A Prong 1: The claim recites a judicial exception.
in which the update process further comprises a regularization update to the second numerical parameters, the regularization update being performed by subtracting a regularization amount from the updated second numerical parameters (mathematical concepts, [0078] “In optional step 205, a regularization amount is subtracted from the new current values of the first and/or second numerical parameters… by removing the amount hλgθ where λ is a regularization multiplier, and gθ = ∇Θ ||∂lG/∂ϕ❘θk,ϕk||2”)
With respect to claim 17:
2A Prong 1: The claim recites a judicial exception.
in which the first and second loss components are both functions of both the first and second numerical parameters (mathematical concepts, in light of specification [0070] “In a general form which covers all these possible loss functions, the overall loss function (objective function) of the discriminator-generator pair can be denoted by: l(θ,ϕ)=[lD(θ,ϕ), lD(θ,ϕ)] (5) [objective function with loss components] The training process amounts to minimizing the loss component lD(θ, ϕ) with respect to θ [1st parameter], and the loss component lG(θ, ϕ) with respect to ϕ[2nd parameter]… In this case, both loss components are functions of both the first parameters θ and the second parameters ϕ.”)
Response to Arguments
Applicant's arguments with respect to the rejection of the claims under 35 U.S.C. 101 have been fully considered but they are not persuasive:
Applicant argues: (p. 9) This rejection should be withdrawn because the claimed invention provides a technical solution to a technical problem. Applicant submits that, even if the claims recite an abstract idea (which is not conceded), the claims are not directed to an abstract idea because any abstract ideas recited the claims are integrated into a practical application…
Examiner answers: In step 2A prong One, the limitation is evaluated if it recites a judicial exception. In step 2A prong Two and step 2B, additional elements are evaluated if those additional elements are integrated into a practical application and are significantly more than the exception.
In step 2A prong One, each of the steps in claim 1 (and also most of all other dependent claims) has corresponding formulas in the specification, therefore those steps are math calculations. In step 2A prong Two and step 2B, the claimed invention does not provide any additional elements that are integrated into a practical application and are significantly more than the exception.
The improvements might be claimed in the limitation “generating respective second or higher order updates to the first and second numerical parameters based on gradients of the loss components with respect to the first and second numerical parameters both for the current values of the first and second numerical parameters and for the first and second intermediate values of the first and second numerical parameters; and.” However, it is an abstract idea itself (as math). If the claim is directed to a judicial exception, it cannot provide an improvement. See MPEP 2106.05(a): “It is important to note, the judicial exception alone cannot provide the improvement.” The claim is directed to an abstract idea.
The applicant is encouraged to amend the claim to overcome 101 rejection.
Applicant argues: (p. 10) The MPEP now explicitly recognizes that an "improved way of training a machine learning model" constitutes an improvement in computer functionality (MPEP4 2106.05(a))… The MPEP states that improvements to system performance based upon "adjustments to parameters of a machine learning model associated with tasks"(MPEP4 2106.05(a)(I)) are "tantamount to how the machine learning model itself would function in operation" (MPEP & 2106.04(d)(1)) and are not subsumed in a mathematical calculation.
Examiner answers: MPEP also refers to Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision), where the steps of the claim in Ex Parte Desjardins may not align with the steps of the claimed invention.
Applicant argues: (p. 11) The claims explicitly recite steps that enable the asserted improvements to the functioning of a computer and to machine learning technology, e.g., claim 1 recites (emphasis added): "A computer-implemented method of training an adversarial machine learning model…
Examiner answers: The feature of “training an adversarial machine learning model… minimize an objective function” is evaluated as an additional element, which is mere instructions to apply an exception – MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception; high level recitation of training a model with an objective function. Further, as mentioned above, all other steps are abstract ideas as mathematical calculations.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Li ("A Novel Generative Model With Bounded-GAN for Reliability Classification of Gear Safety" 20190103) teaches (Li, p. 5, 3) Network Optimization "The optimizer computes exponential moving averages of gradients {∇(D), ∇(G)} and Hessian matrices {H(D), H(G)} [second or higher order updates to the first and second numerical parameters D and G]… The estimations of first-moment gradient and Hessian matrices"). However, Li does not teach “generating respective second or higher order updates… both for the current values of the first and second numerical parameters and for the first and second intermediate values of the first and second numerical parameters; and” (see spec [0076]).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SU-TING CHUANG whose telephone number is (408)918-7519. The examiner can normally be reached Monday - Thursday 8-5 PT.
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, Usmaan Saeed can be reached at (571) 272-4046. 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.
/SU-TING CHUANG/Examiner, Art Unit 2146
Prosecution Insights