DETAILED ACTION
This action is in response to the application filed 4/11/2022. Claims 1-21 are pending and have been examined.
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 1/29/2026 has been entered.
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 .
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-21 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 1, as amended, recites “wherein the compressed residual global weight update … is a residual of the compressed residual local weight update” and “wherein the compressed residual local weight update … is a respective residual of another respective compressed residual global weight update”. While the instant specification discloses calculating a global weight update based on local weight updates constructed from residual local weight updates and calculating a local weight update based on a global weight update constructed from a residual global weight update in Framework 1 ([0081]-[0088]), it does not disclose a compressed residual global weight update that is a residual of a compressed residual local weight update, or a compressed residual local weight update that’s a residual of a compressed residual global weight update. Thus, these limitations contain new matter not present in the initial disclosure. This new matter is present in substantially similar independent claims 8 and 15, and is inherited by all dependent claims.
In light of the instant specification (particularly paragraphs [0076] and [0083-0084]), in addition to “residual” being a known term of art, a “residual” is interpreted as the difference between a real / expected / true / intended value and a predicted / approximated / estimated value. Thus, the Examiner interprets “wherein the compressed residual global weight update … is a residual of the compressed residual local weight update” as stating that the compressed residual global weight update is equal to the difference between a predicted compressed residual local weight update and a true compressed residual local weight update. Similarly, the Examiner interprets “wherein the compressed residual local weight update … is a respective residual of another respective compressed residual global weight update” as stating that the compressed residual local weight update is equal to the difference between a predicted compressed residual global weight update and a true compressed residual global weight update.
Response to Arguments
The following responses address arguments and remarks made in the instant remarks dated 1/29/2026
Objections
In light of the instant amendments, previous objections to the claims have been withdrawn.
In the Advisory Action mailed 1/27/2026, the Examiner stated that the proposed amendments dated 12/18/2025 would incur objections to the claims. These objections have not been made in light of the revised instant amendments.
112 Rejections
On pages 13-15 of the instant remarks, the Applicant argues that the amended claims overcome possible 112(b) rejections noted in a prior Advisory Action:
“The Examiner further introduced new rejections under 35 U.S.C. 112(b) as follows
…
Accordingly, claim 1 is amended
…
Independent claim 1 is from the perspective of the server, and is based on Framework 1 as
shown in Table 2 of the filed specification. Independent claim 8 has been amended similarly from
the perspective of the institute/client, and is based on Framework 1 as shown in Table 2 of the filed
specification. Independent claim 15 has been amended similarly from the perspective of the server,
and is based on Framework 2 as shown in Table 3 of the filed specification
…
Based on the foregoing, it is respectfully submitted that the claims as currently amended
are sufficiently clear under 35 U.S.C. 112(b).”
The Applicant’s arguments above are persuasive. In the Advisory Action mailed 1/27/2026, the Examiner stated that the proposed amendments dated 12/18/2025 would incur claim rejections under 35 U.S.C. 112(b). The deficiencies that would have incurred these rejections are not present in the instant amendments, so the proposed rejections are not made.
New rejections under 35 U.S.C. 112(a) have been made in light of the instant amendments.
103 Rejections / Allowable Subject Matter
In light of the instant amendments, previous rejections under 35 U.S.C. 103 for claims 1-21 have been withdrawn. The Examiner notes that claims 1-21 are found to be allowable over the prior art under 35 U.S.C. 102 and 103 only under the Examiner’s specific interpretation of “residual” in the amended limitations, as discussed in the 112(a) rejections section. Said interpretation is reiterated below.
In light of the instant specification (particularly paragraphs [0076] and [0083-0084]), in addition to “residual” being a known term of art, a “residual” is interpreted as the difference between a real / expected / true / intended value and a predicted / approximated / estimated value. Thus, the Examiner interprets “wherein the compressed residual global weight update … is a residual of the compressed residual local weight update” as stating that the compressed residual global weight update is equal to the difference between a predicted compressed residual local weight update and a true compressed residual local weight update. Similarly, the Examiner interprets “wherein the compressed residual local weight update … is a respective residual of another respective compressed residual global weight update” as stating that the compressed residual local weight update is equal to the difference between a predicted compressed residual global weight update and a true compressed residual global weight update.
Allowable Subject Matter
Claims 1-21 would be allowed if not for their current rejections under 35 U.S.C. 112(a).
Regarding claim 1, an apparatus, wherein the compressed residual global weight update … is a residual of the compressed residual local weight update … ; wherein the compressed residual local weight update … is a respective residual of another respective compressed residual global weight update, is not taught by the prior art of record. The closest prior arts of record are Samek et al. (CONCEPTS FOR DISTRIBUTED LEARNING OF NEURAL NETWORKS AND/OR TRANSMISSION OF PARAMETERIZATION UPDATES THEREFOR, published 3/4/2021, US 2021/0065002 A1) and Spratling (A review of predictive coding algorithms, 2017, Brain and Cognition 112 92–97)
Samek discloses an apparatus for transmitting compressed local and global weight updates between servers clients in a federated learning system: “The goal in distributed training is to train a global model, using all of the clients training data, without sending around this data. This is achieved by performing the following steps: Clients that want to contribute to the global training first synchronize with the current global model, by downloading 32 it from a server. They then compute 34 a local weight-update using their own local data and upload 36 it to the server. At the server all weight-updates are aggregated 38 to form a new global model.” (Samek, [0067])
Samek does not disclose calculating or utilizing residuals of updates.
Spratling discloses a method of calculating residuals of time-series data values and transmitting / receiving them in lieu of the original values:
“Digital signal processing concerns the manipulation and analysis of a continuous signal, x, sampled at discrete time points (indexed by i) so that the signal is represented as a sequence of numbers, x(i), called a ‘time series’” (Spratling, page 93, left column, paragraph 3);
“e is used to denote the error between the reconstruction and the actual sensory input (or the ‘residual’)” (Spratling, page 93, left column, paragraph 20)
“the estimated value of the signal, as calculated by Eq. (1), is subtracted from the true value, x(i), to determine the residual error, e(i), for transmission:
PNG
media_image1.png
77
287
media_image1.png
Greyscale
This residual has a smaller dynamic range than the original signal, and hence, can be transmitted with greater accuracy using the same bandwidth” (Spratling, page 93, right column, paragraph 2); “it should be noted that if only the residual error is transmitted, then the receiver … cannot recover the components of the signal that have been removed” (Spratling, page 93, right column, paragraph 3).
Spratling does not disclose calculating or utilizing the residual of a residual value.
Therefore, the prior art of record, individually or in combination, does not disclose the entirety of claim 1 as a whole. Similar reasoning applies to substantially similar independent claims 8 and 15. Claims 2-7, 9-14, and 16-21 would be allowable at least due to their dependence on the independent claims.
The Examiner notes that claims 1-21 are found to be allowable over the prior art under 35 U.S.C. 102 and 103 only under the Examiner’s specific interpretation of “residual” in the amended limitations discussed previously in the 112(a) rejections section. Said interpretation is reiterated below.
In light of the instant specification (particularly paragraphs [0076] and [0083-0084]), in addition to “residual” being a known term of art, a “residual” is interpreted as the difference between a real / expected / true / intended value and a predicted / approximated / estimated value. Thus, the Examiner interprets “wherein the compressed residual global weight update … is a residual of the compressed residual local weight update” as stating that the compressed residual global weight update is equal to the difference between a predicted compressed residual local weight update and a true compressed residual local weight update. Similarly, the Examiner interprets “wherein the compressed residual local weight update … is a respective residual of another respective compressed residual global weight update” as stating that the compressed residual local weight update is equal to the difference between a predicted compressed residual global weight update and a true compressed residual global weight update.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Liu et al. (A Double Residual Compression Algorithm for Efficient Distributed Learning, published 2019, arXiv:1910.07561v1) discloses a method of transmitting local and global model updates in a distributed learning system via residuals
Thapa et al. (SplitFed: When Federated Learning Meets Split Learning, 2020, arXiv:2004.12088v2) teaches a federated learning system where local weight updates are calculated across multiple machines
Chen et al. (A new lossy compression algorithm for wireless sensor networks using Bayesian predictive coding, 2020, Wireless Netw 26, 5981–5995 (2020). https://doi.org/10.1007/s11276-020-02425-w) teaches a method of using predictive coding to encode data being transmitted over a network
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Aaron P Gormley whose telephone number is (571)272-1372. The examiner can normally be reached Monday - Friday 12:00 PM - 8:00 PM EST.
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, Michelle T Bechtold can be reached at (571) 431-0762. 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.
/AG/Examiner, Art Unit 2148
/MICHELLE T BECHTOLD/Supervisory Patent Examiner, Art Unit 2148