CTNF 18/957,019 CTNF 84252 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. Detailed Action 07-42-04 AIA 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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 3/16/26 has been entered. In amendments dated 3/16/26, Applicant amended claims 1-2, 6, 9, 11, 16, 18, and 20-21, canceled no claims, and added no new claims. Claims 1-9 and 11-21 are presented for examination. Applicant is advised that the instant application is now being examined by Examiner Bruce Moser. Objections Claim 1 is objected to because of the following informality: each of the second and third limitations recite “a second hash value,” making the antecedent basis of the second reciting unclear. Rejections under 35 U.S.C. 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 1-9 and 1-21 are rejected under 35 U.S.C. 101 because the claimed invention is directed to mental processes without significantly more. Independent claim 1 recites based on the modified version of the data chunk, generating a second hash value different from the first hash value, generating a second version of the manifest that references the modified version of the data chunk with a second hash value comprising an entry comprised in the manifest, and based on a comparison of the first version of the manifest and the second version of the manifest, generating data corresponding to a snapshot differences report for a snapshot of changes to the tensor over time, wherein the snapshot differences report is representative of a difference between first text of the first hash value and second text of the second hash value. Generating a hash value and generating a second hash value are conventional math and recited broadly and a mental process accomplishable in the human mind or on paper, generating a second version of a manifest, generating data corresponding to snapshot differences, and comparing manifest versions are each recited broadly and are menta processes accomplishable in the human mind or on paper. The claim recites an additional element of receiving, from a model training engine, a modified version of a data chunk of stored data chunks, wherein the stored data chunks comprise a tensor, and wherein the stored data chunks are referenced by a first version of a manifest that identifies a most recently stored version of the data chunk with a corresponding first hash value, which is a data gathering step and insignificant extra-solution activity. The claim recites at least one memory and at least one processor which are generic components of a computer. Examiner notes specification paragraphs 0001 and 0023 state that “both new and older versions of a tensor may need to be persisted for future use” for training AI models and paragraph 0023 also mentions “Problems resulting from inefficient storage of training data may be aggravated as the use of raw data that includes complex multimedia continues to increase.” Paragraph 0024 states “One or more embodiments [of the instant invention] provide a tensor storage that can take tensor snapshots and track differences between snapshots so that only differences in a tensor are stored [emphasis added].” Examiner notes claim 1 covers this embodiment in broad terms but the claim steps do not recite a particular improvement in any technology or function of a computer per MPEP 2106.04(d) and do not recite any unconventional steps in the invention per MPEP 2106.05(a). Therefore, the recited mental processes are not integrated into a practical application. Taking the claim as a whole, the data gathering step is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. The at least one memory and at least one processor are still generic components of a computer. Thus the claims does not include additional elements that are sufficient to amount to significantly more than the recited mental processes. Independent claim 11 recites manipulating, by the system, a cached copy of the data chunks to change a copy of the chunk from the first version of the chunk to a second version of the chunk, different from the first version of the chunk; and generating a second version of the text manifest that references the second version of the chunk with a second hash value different from the first hash value; and based on a comparison of the first version of the text manifest and the second version of the text manifest, generating data corresponding to a snapshot differences report representative of a difference between the first hash value and the second hash value. Manipulating a cached copy of data chunks to change a copy to a different version is recited broadly and is a mental process accomplishable in the human mind or on paper. Generating a second version of the manifest, comparing versions of the manifest, and generating data corresponding to a snapshot differences report are each recited broadly and is a mental process accomplishable in the human mind or on paper. This claim recites additional elements of storing, by a system comprising at least one processor, a tensor as data chunks in a storage device, wherein a chunk of the data chunks comprises a first version of the chunk, wherein a first version of a text manifest references the first version of the chunk with a corresponding first hash value; and copying, by the system, to the storage device, the second version of the chunk to the data chunks, which are each storing steps and insignificant extra-solution activity. Examiner notes the problems discussed in specification paragraphs 0001 and 0023 and the improvements in paragraph 0024 as noted above. Examiner also notes the claim steps do not recite a particular improvement in any technology or function of a computer per MPEP 2106.04(d) and do not recite any unconventional steps in the invention per MPEP 2106.05(a). Therefore, the recited mental processes are not integrated into a practical application. Taking the claim as a whole, the storing steps are storing data in a memory and are routine and conventional activities per the list of such activities in MPEP 2106.05(d) part II. Thus the claims does not include additional elements that are sufficient to amount to significantly more than the recited mental processes. Independent claim 16 recites updating a text snapshot record that references a first storage location of the data block with a first hash value to further include a second storage location of the modified data block with a second hash value different than the first hash value, resulting in an updated text snapshot record, and updating a text record with a has value is recited broadly and a mental process accomplishable in the human mind or on paper. This claim recites an additional element of receiving a modified version of a data block of stored data blocks, wherein the stored data blocks store portions of a multidimensional array of model training weights representative of a tensor, which is a data gathering step and insignificant extra-solution activity. The claim recites a non-transitory machine-readable medium comprising executable instructions which is a generic component of a computer. Examiner notes the problems discussed in specification paragraphs 0001 and 0023 and the improvements in paragraph 0024 as noted above. Examiner also notes the claim steps do not recite a particular improvement in any technology or function of a computer per MPEP 2106.04(d) and do not recite any unconventional steps in the invention per MPEP 2106.05(a). Therefore, the recited mental processes are not integrated into a practical application. Taking the claim as a whole, the data gathering step is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. The non-transitory machine-readable medium comprising executable instructions is still a generic component of a computer. Thus the claims does not include additional elements that are sufficient to amount to significantly more than the recited mental processes. Claim 2 recites receiving a request to input the tensor to the model training engine, and receiving a request is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II, based on the second version of the text manifest, selecting the most recently stored version of the data chunk, and selecting a version involves evaluating and is a mental process, and based on the most recently stored version of the data chunk, inputting the tensor to the model training engine, and inputting data is recited broadly and is a mental process accomplishable in the human mind or on paper. Claim 3 recites wherein the modified version of the data chunk of stored data chunks is received from the training engine based on a checkpoint operation, and receiving a version of data chunk is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 4 recites wherein the request to input the tensor to the model training engine comprises a request to input data from the checkpoint operation, and receiving a request to input data is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 5 recites wherein the second version of the text manifest further references a stored version of another chunk of the stored data chunks as a most recently created version of the other chunk, and storing data is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 6 recites wherein the first version of the text manifest and the second version of the text manifest are respectively comprised in a first file system file of a file system and a second file system file of the file system, and comprising versions of the manifest is storing data which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 7 recites wherein the second file system file comprises entries referencing the data chunks of the tensor, and wherein the entries comprise references to most recent version of the data chunk, and comprising versions of the manifest is storing data which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 8 recites wherein the data chunks, the first file system file, and the second file system file are stored in a directory of the file system, and storing data is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 9 recites based on the second version of the text manifest, generating version tracking data for the tensor, and generating tracking data is recited broadly and is a mental process accomplishable in the human mind or on paper. Claim 12 recites before the copying of the second version of the chunk, based on the snapshot differences report, identifying, by the system, that the second version of the chunk is different from the first version of the chunk, and identifying a chunk version is different is evaluating and a mental process, and copying, by the system, the second version of the chunk based on the identifying, which is storing data and is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 13 recites receiving, by the system, a request to perform a checkpoint operation on a process of manipulating the data chunks, wherein copying the second version to the data chunks comprises the checkpoint operation, which comprises the change to the copy of the chunk, and receiving a request to perform a checkpoint operation is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 14 recites wherein the system comprises a model training system that utilizes tensors to train a machine learning model, and utilizing tensors to train a machine learning model is a conventional activity or a machine learning model and not more than a mental process per Recentive Analytics v. Fox Broadcasting Corp. (134 F.4th 1205, 2025 U.S.P.Q.2d 628) . Claim 15 recites receiving, by the system, from a raw data processing engine, the tensor, wherein the tensor was generated by a transformation of raw data, and receiving tensor data is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 17 recites wherein the receiving of the modified version of the data block comprises receiving the modified version from a training backend that generated the modified version by transforming source data of the multidimensional array to modify a portion of the multidimensional array corresponding to the data block, and receiving a data block is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 18 recites comparing the text snapshot record to the updated text snapshot record to identify the modified version of the data block, resulting in a snapshot update of the multidimensional array, and comparing snapshot records involves evaluating and is a mental process; and communicating, to a compute node that uses the multidimensional array as training data, the snapshot update of the multidimensional array, and communicating an update is recited broadly and amounts to sending data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 19 recites wherein the receiving of the modified version of the data block comprises receiving the modified version from a compute node that generated the modified version of the data block based on a modification of a model weight stored in the multidimensional array, and receiving a modified version is recited broadly and amounts to receiving data across a network per specification figure 1 191, figure 2 290, and paragraphs 0025 and 0038, which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Claim 20 recites analyzing the text snapshot record and the updated text snapshot record, and analyzing a record is recited broadly and is a mental process accomplishable in the human mind or on paper; and based on the analyzing, generating a version tracking report for the multidimensional array, wherein the version tracking report identifies that second hash value is different from the first hash value, and generating a report is recited broadly and is a mental process accomplishable in the human mind or on paper. Claim 21 recites wherein the text snapshot record and the updated text snapshot record are respectively comprised in a first file system file of a file system and a second file system file of the file system, and comprising text records in a file system is storing the text records which is routine and conventional activity per the list of such activities in MPEP 2106.05(d) part II. Rejections under 35 U.S.C. 102 07-07-aia AIA 07-07 The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action: A person shall be entitled to a patent unless – 07-08-aia AIA (a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention. 07-15 AIA Claim s 1-4, 6-9, and 11-21 are rejected under 35 U.S.C. 102( a)(1 ) as being anticipated by Gangumalla et al (US 20250342179), hereafter Gangumalla . With respect to claim 1, Gangumalla teaches: receiving, from a model training engine, a modified version of a data chunk of stored data chunks, wherein the stored data chunks comprise a tensor, and wherein the stored data chunks are referenced by a first version of a manifest that identifies a most recently stored version of the data chunk with a corresponding first hash value (paragraphs 0017, 0020 input data sets for training AI models (tensors), paragraph 0018 documents or portions of documents modified in input data set, paragraph 0037 making chunk-level updates, checksums of the chunks, paragraph 0010 chunk updates in an index (manifest)), based on the modified version of the data chunk, generating a second hash value different from the first hash value (paragraph 0035 generate checksums of modified chunks, different from checksums of unmodified chunks), generating a second version of the manifest that references the modified version of the data chunk with a second hash value comprising an entry comprised in the manifest (paragraph 0010 index updated (second version) with updated chunks, also paragraphs 0018-0019), and based on a comparison of the first version of the manifest and the second version of the manifest, generating data corresponding to a snapshot differences report for a snapshot of changes to the tensor over time, wherein the snapshot differences report is representative of a difference between first text of the first hash value and second text of the second hash value (paragraph 0031 analyze index for IDs of changed documents, updates index for snapdiff report, paragraph 0035 compare checksums of versions of chunks for report). With respect to claim 1, Gangumalla teaches at least one memory and at least one processor (paragraphs 0088 figure 8 memory 813, processor 812). With respect to claim 11, Gangumalla teaches: storing, by a system comprising at least one processor, a tensor as data chunks in a storage device, wherein a chunk of the data chunks comprises a first version of the chunk, wherein a first version of a text manifest references the first version of the chunk with a corresponding first hash value (paragraphs 0017, 0020 input data sets for training AI models (tensors), paragraph 0018 documents or portions of documents modified in input data set, paragraph 0037 making chunk-level updates, checksums of the chunks, paragraph 0010 chunk updates in an index (manifest)); manipulating, by the system, a cached copy of the data chunks to change a copy of the chunk from the first version of the chunk to a second version of the chunk, different from the first version of the chunk (paragraphs 0063, 0088 data for chunks etc. stored in cache, paragraph 0035 generate checksums of modified chunks, different from checksums of unmodified chunks); and generating a second version of the text manifest that references the second version of the chunk with a second hash value different from the first hash value (paragraph 0010 index updated (second version) with updated chunks, also paragraphs 0018-0019); based on a comparison of the first version of the text manifest and the second version of the text manifest, generating data corresponding to a snapshot differences report representative of a difference between the first hash value and the second hash value (paragraph 0031 analyze index for IDs of changed documents, updates index for snapdiff report, paragraph 0035 compare checksums of versions of chunks for report); and copying, by the system, to the storage device, the second version of the chunk to the data chunks (paragraph 0035 modification of chunks in stored documents). With respect to claim 16, Gangumalla teaches: receiving a modified version of a data block of stored data blocks, wherein the stored data blocks store portions of a multidimensional array of model training weights representative of a tensor (paragraph 0037 making chunk-level updates, checksums of the chunks, also paragraph 0020 input data as multidimensional arrays); and updating a text snapshot record that references a first storage location of the data block with a first hash value to further include a second storage location of the modified data block with a second hash value different than the first hash value, resulting in an updated text snapshot record (paragraph 0010 update the index, paragraph 0043 checksums of modified chunks stored in index, paragraph 0030 generating updated snapshots at different times after detecting changes to input data set). With respect to claim 16, Gangumalla teaches a non-transitory machine-readable medium (paragraph 0063 instructions stored on non-transitory computer-readable media). With respect to claim 2, Gangumalla teaches: receiving a request to input the tensor to the model training engine (paragraph 0017 inputting input set for training AI model), based on the second version of the text manifest, selecting the most recently stored version of the data chunk (paragraph 0017 selecting updated version of input set), and based on the most recently stored version of the data chunk, inputting the tensor to the model training engine (paragraph 0017 retraining AI model with updated input set). With respect to claim 3, Gangumalla teaches wherein the modified version of the data chunk of stored data chunks is received from the training engine based on a checkpoint operation (paragraph 0029 updated changes copied to input set as a checkpoint operation per specification paragraph 0004). With respect to claim 4, Gangumalla teaches wherein the request to input the tensor to the model training engine comprises a request to input data from the checkpoint operation (paragraph 0029 updated changes copied to input set as a checkpoint operation per specification paragraph 0004). With respect to claim 6, Gangumalla teaches wherein the first version of the text manifest and the second version of the text manifest are respectively comprised in a first file system file of a file system and a second file system file of the file system (paragraph 0019 using snapshots from plural storage systems, paragraph 0024 distributed storage systems). With respect to claim 7, Gangumalla teaches wherein the second file system file comprises entries referencing the data chunks of the tensor, and wherein the entries comprise references to most recent version of the data chunk (paragraph 0019 snapshot references chunks). With respect to claim 8, Gangumalla teaches wherein the data chunks, the first file system file, and the second file system file are stored in a directory of the file system (paragraph 0010 index updated, stored in a directory of the storage system). With respect to claim 9, Gangumalla teaches wherein the operations further comprise, based on the second version of the text manifest, generating version tracking data for the tensor (paragraph 0010 index keeping track of versions). With respect to claim 12, Gangumalla teaches: before the copying of the second version of the chunk, based on the snapshot differences report, identifying, by the system, that the second version of the chunk is different from the first version of the chunk (paragraphs 0030, 0031 identify different versions of a chunk), and copying, by the system, the second version of the chunk based on the identifying (paragraphs 0030, 0031 updates document in index based on identified differences). With respect to claim 13, Gangumalla teaches receiving, by the system, a request to perform a checkpoint operation on a process of manipulating the data chunks, wherein copying the second version to the data chunks comprises the checkpoint operation, which comprises the change to the copy of the chunk (paragraph 0029 updated changes copied to input set as a checkpoint operation per specification paragraph 0004). With respect to clam 14, Gangumalla teaches wherein the system comprises a model training system that utilizes tensors to train a machine learning model (paragraph 0017 system for training AI model with input sets). With respect to claim 15, Gangumalla teaches receiving, by the system, from a raw data processing engine, the tensor, wherein the tensor was generated by a transformation of raw data (paragraph 0017 receive input set to train AI model). With respect to claim 17, Gangumalla teaches wherein the receiving of the modified version of the data block comprises receiving the modified version from a training backend that generated the modified version by transforming source data of the multidimensional array to modify a portion of the multidimensional array corresponding to the data block (paragraphs 0017, 0025 receiving training data, paragraph 0020 input set is multi-dimensional data). With respect to claim 18, Gangumalla teaches: comparing the text snapshot record to the updated text snapshot record to identify the modified version of the data block, resulting in a snapshot update of the multidimensional array (paragraph 0030 comparing snapshots records in snapdiff processor to detect changes); and communicating, to a compute node that uses the multidimensional array as training data, the snapshot update of the multidimensional array (paragraph 0030 communicate differences to snapdiff process). With respect to claim 19, Gangumalla teaches wherein the receiving of the modified version of the data block comprises receiving the modified version from a compute node that generated the modified version of the data block based on a modification of a model weight stored in the multidimensional array (paragraph 0034 changes in tuning parameters (weights) for input data sets). With respect to claim 20, Gangumalla teaches: analyzing the text snapshot record and the updated text snapshot record (paragraph 0030 analyze snapshots at T1, T2); and based on the analyzing, generating a version tracking report for the multidimensional array, wherein the version tracking report identifies that second hash value is different from the first hash value (paragraph 0030 snapdiff report based on analyzing). With respect to claim 21, Gangumalla teaches wherein the text snapshot record and the updated text snapshot record are respectively comprised in a first file system file of a file system and a second file system file of the file system (paragraph 0019 using snapshots from plural storage systems, paragraph 0024 distributed storage systems). Responses to Applicant’s Remarks With respect to objection to claim 11, in view of Applicant’s amendments, this objection is withdrawn. With respect to rejections of claims 1-9, 11-17, and 19-21 under 35 U.S.C. 103 by Potnis, Tensor paper, and in further view of Mikar, Applicant’s amendments overcome Potnis’ Tensor’s and Mikar’s teachings, in particular generating a first and second hash values for the data chunks. Examiner conducted another search of the prior art and found Gangumalla, which Examiner believes teaches claims 1-4, 6-9, and 11-21 as shown above. Inquiry Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRUCE M MOSER whose telephone number is (571)270-1718. The examiner can normally be reached M-F 9a-5p. 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, Boris Gorney can be reached at 571 270-5626. 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. /BRUCE M MOSER/Primary Examiner, Art Unit 2154 6/10/26 Application/Control Number: 18/957,019 Page 2 Art Unit: 2154 Application/Control Number: 18/957,019 Page 3 Art Unit: 2154 Application/Control Number: 18/957,019 Page 4 Art Unit: 2154 Application/Control Number: 18/957,019 Page 5 Art Unit: 2154 Application/Control Number: 18/957,019 Page 6 Art Unit: 2154 Application/Control Number: 18/957,019 Page 7 Art Unit: 2154 Application/Control Number: 18/957,019 Page 8 Art Unit: 2154 Application/Control Number: 18/957,019 Page 9 Art Unit: 2154 Application/Control Number: 18/957,019 Page 10 Art Unit: 2154 Application/Control Number: 18/957,019 Page 11 Art Unit: 2154 Application/Control Number: 18/957,019 Page 12 Art Unit: 2154 Application/Control Number: 18/957,019 Page 13 Art Unit: 2154 Application/Control Number: 18/957,019 Page 14 Art Unit: 2154 Application/Control Number: 18/957,019 Page 15 Art Unit: 2154
Prosecution Insights