DETAILED ACTION
This action is in response to the filing of 5-11-2026. Claims 1-20 are pending and have been considered below:
Allowable Subject Matter
Claim 12 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim Rejections - 35 USC § 103
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 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.
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.
Claims 1, 8 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over “Review of second-order optimization technique in artificial neural network backpropagation” Tan et al. (“Tan”) 2019, Pages 1-8 in view of Haidar et al. (“Haidar” 20200134415 A1) and Network Quantization with Element-wise Gradient Scaling, Junghyup Lee et al., 4-2-2021 (“Lee”) pages 1-10.
Claim 1: Tan discloses a method comprising: approximating, to a second-order accuracy (Section 1, page 2), an approximation of a gradient of a parameter (Section 1, page 2 “The Hessian-free method was then proposed as an alternative second-order optimization technique with the help of conjugate gradient algorithm.” And Table 1: Hessian-free)
adjusting the parameter based on the approximation of the gradient resulting in an adjusted parameter; and operating the NN using the adjusted parameter (Section 3; fine tuning of parameter).
It is generally understood that a Deep neural network as found in Tan (Figure 1) would have a latent space. However, to more clearly capture using a non-differentiable sample operation, and a discrete latent variable of a neural network (NN), Haidar is disclosed. Haidar provides a Deep neural network (GAN) which provides a latent representation from the network (Paragraph 29).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to use a known technique to improve a similar device and provide clear latent expressions utilizing the DNN of Tan. One would have been motivated to provide the functionality for improved dimensionality, for finding hidden characteristics.
Lee is also provided because it discloses a second order functionality that determines gradient with discrete values (Page 2, Column 1-2, Paragraph 2).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to use a known technique to improve a similar device and provide discrete variables for backpropagation in Tan. One would have been motivated to provide the functionality for improved dimensionality, for finding hidden characteristics.
Claims 8 and 17 are similar in scope to claim 1 and therefore rejected under the same rationale.
Claim 8 non-transitory machine-readable medium (Haidar: Paragraphs 50 and 68)
Claim 17 memory and processor (Haidar: Paragraphs 50 and 68)
Claims 2-3, 5-7, 9-11, 14-16 and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over “Review of second-order optimization technique in artificial neural network backpropagation” Tan et al. “Tan” 2019, Pages 1-8, Haidar et al. (“Haidar” 20200134415 A1) and Network Quantization with Element-wise Gradient Scaling, Junghyup Lee et al., 4-2-2021 (“Lee”) pages 1-10 in further view of Ye at al. (“Ye” 20230259779 A1).
Claim 2: Tan, Haidar and Lee disclose a method of claim 1, wherein determining the approximation of the gradient includes: sampling a one hot encoding of output of the NN resulting in a sample (Haidar: Paragraph 45; one-hot representations); however may not explicitly disclose computing a first combination of the sample and a tempered probability distribution of outcomes for the output of the NN; computing a second probability distribution of outcomes based on the tempered probability distribution and the output of the NN; computing a second combination of the probability distribution of outcomes and the second probability distribution of outcomes resulting in a third probability distribution of outcomes; altering a value of the sample based on the third probability distribution of outcomes resulting in an altered value; and
PNG
media_image1.png
5
3
media_image1.png
Greyscale
wherein adjusting the parameter is based on the altered value.
Ye discloses a sample from a one-hot feature (Paragraph 83), further Ye discloses a continuous gumbel-softmax distribution (Paragraph 79) and further provides reparameterization for a distribution (Paragraph 81). Hence the system provides multiple distributions where the values are altered (using continuous Gumbel-Softmax distribution). Therefore, it would have been obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to use a known technique to improve a similar device and provide the continuous Gumbel-Softmax for multiple distributions in the modified Tan. One would have been motivated to provide the functionality for improved sampling and enhanced analysis for optimization.
Claim 3: Tan, Haidar, Lee and Ye disclose a method of claim 2, wherein determining the approximation of the gradient further comprises: determining a first probability distribution of outcomes based on output of the NN; and determining, based on the first probability distribution of outcomes, a one hot encoding (Ye: Paragraph 83; probability distribution).
Claim 5: Tan, Haidar, Lee and Ye disclose a method of claim 2, wherein the second combination is a weighted difference between the probability distribution of outcomes and the second probability distribution of outcomes (Ye: Paragraph 79; the continuous distribution would provide differences).
Claim 6: Tan, Haidar, Lee and Ye disclose a method of claim 2, wherein a temperature of the tempered probability distribution is greater than, or equal to, one (Ye: Paragraph 83; non-zero temperature can be greater than 1) could be greater than 1.
Claim 7: Tan, Haidar, Lee and Ye disclose a method of claim 2, wherein the operations are constrained to a baseline subtraction that is set to an expected value of the sample (Ye: Paragraph 124; default weightings used to differentiate the loss (subtraction)).
Claims 9-10 and 18-19 similar in scope to claim 2 and therefore rejected under the same rationale.
Claims 11 and 20 are similar in scope to claim 3 and therefore rejected under the same rationale.
Claim 14 is similar in scope to claim 5 and therefore rejected under the same rationale.
Claim 15 is similar in scope to claim 6 and therefore rejected under the same rationale.
Claim 16 is similar in scope to claim 7 and therefore rejected under the same rationale.
Claims 4 and 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over “Review of second-order optimization technique in artificial neural network backpropagation” Tan et al. “Tan” 2019, Pages 1-8, Haidar et al. (“Haidar” 20200134415 A1), Ye at al. (“Ye” 20230259779 A1) and Network Quantization with Element-wise Gradient Scaling, Junghyup Lee et al., 4-2-2021 (“Lee”) pages 1-10 in further view of “Average-tempered stable subordinate with applications”, “Xia”, abstract 6-16-2021.
Claim 4: Tan, Haidar, Lee and Ye disclose a method of claim 2, but may not explicitly disclose wherein the first combination is an average. Xia as provided discloses tempered distributions with an averaging process (abstract).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to use a known technique to improve a similar device and provide averaged distributions in the modified Tan. One would have been motivated to provide the averaging functionality for improved representation and normalized analysis.
Claim 13 is similar in scope to claim 4 and therefore rejected under the same rationale.
Response to Arguments
Applicants’ arguments have been considered but are moot because of the new grounds of rejection incorporating Lee. Further, upon review there is no explicit language of a continuous vs discrete latent variable in Tan. It is further examiners’ understanding that continuous variables can take any value. Additionally, the discrete functionality is not captured by relying on Tan. Tan also does not provide any negative recitations against discrete variables even if it is implied that continuous variables are utilized. Therefore, the additional references of Haidar and Lee which provide second order functionality and discrete variables could be incorporated to provide the functionality under BRI.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
20200394559 A1 ZHANG ET AL. ABSTRACT
Applicants’ amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHERROD L KEATON whose telephone number is (571)270-1697. The examiner can normally be reached on MONDAY -FRIDAY 9:30-5.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Michelle Bechtold can be reached on 571-272-4124. The fax phone number for the organization where this application or proceeding is assigned is 571-273-3800.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/SHERROD L KEATON/Primary Examiner, Art Unit 2148
7-28-2026
Prosecution Insights