DETAILED ACTION
This action is written in response to the application filed 1/18/24. 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 § 101
Claims 16-30 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. 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.
In determining whether the claims are subject matter eligible, the Examiner applies the 2019 USPTO Patent Eligibility Guidelines, as well as guidance from MPEP § 2106.
Step 1: Is the claim to a process, machine, manufacture, or composition of matter? Yes—claim 16 recites a method, which is a process.
Step 2A, prong one: Does the claim recite an abstract idea, law of nature or natural phenomenon? Yes—the claim recites one or more limitations which—under their broadest reasonable interpretation—covers performance of the limitation in the mind (see table below).
Claim limitation
Examiner analysis
Claim 16: A computer-implemented method of determining a quantization range for tensors of an artificial neural network, the method comprising:
observing a saturation ratio at a current iteration from the tensors and a quantization range of the artificial neural network; and
This is a mental process akin to a human evaluation/observation.
adjusting the quantization range such that the observed saturation ratio follows a preset target saturation ratio.
This is a mental process akin to a human evaluation/judgment.
Because the claim recites limitations which can practically be implemented as mental processes, the claim recites a mental process.
Step 2A, prong two: Does the claim recite additional elements that integrate the judicial exception into a practical application? No—the claim does not recite even generic computer hardware.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No. The only limitation on the performance of the described method is that it be “computer-implemented”. The claim thus recites computing components only at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components. The statement that the method is performed by computer does not satisfy the test of “inventive concept.” See Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 573 U.S. 208, 134 S. Ct. 2347, 2360 (2014).
For the reasons above, claim 16 is rejected as being directed to non-patentable subject matter under §101. This rejection applies equally to independent claims 24-26 and 29-30, which each recite a related method/system/computer-readable medium, as well as to all pending dependent claims. The additional limitations of the dependent claims are addressed briefly below. Taken alone, the additional elements of the dependent claims above do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation.
Claim limitation
Examiner analysis
17. The method of claim 16, wherein the observing of the saturation ratio comprises calculating the ratio of the number of tensors outside the quantization range to the number of tensors.
This is a mental process akin to a human evaluation/judgment.
18. The method of claim 16, wherein the adjusting of the quantization range comprises:
calculating a current moving average based on the observed saturation ratio and a past moving average calculated from saturation ratios observed at previous iterations; and
adjusting the quantization range based on a difference between the current moving average and the target saturation ratio.
This is a mental process akin to a human evaluation/judgment.
This is a mental process akin to a human evaluation/judgment.
19. The method of claim 18, wherein the calculating of the current moving average comprises calculating the current moving average through a weighted sum of the past moving average and the observed saturation ratio.
This is a mental process akin to a human evaluation/judgment.
20. The method of claim 19, further comprising adjusting a weight of the past moving average and a weight of the observed saturation ratio.
This is a mental process akin to a human evaluation/judgment.
21. The method of claim 18, wherein the adjusting of the quantization range comprises:
calculating an amount of change in the quantization range based on the difference between the current moving average and the target saturation ratio; and
adjusting the quantization range according to the amount of change in the quantization range.
This is a mental process akin to a human evaluation/judgment.
This is a mental process akin to a human evaluation/judgment.
22. The method of claim 16, further comprising setting an initial value of the quantization range based on batch normalization parameters of the artificial neural network.
This is a mental process akin to a human evaluation/judgment.
23. The method of claim 16, wherein the tensors are derived from either training data in a training stage of the artificial neural network or user data in an inference stage.
This is a mental process akin to a human evaluation/judgment.
Claim Interpretation - 35 USC § 112(f)
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. - An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f):
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f). The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f), is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f). The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f), is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f), except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f), except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f), because the claim limitations uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are:
In claim 24: “an observer configured to …” and “a controller configured to …”.
In claim 30: “a range determination unit configured to …” and “a quantization unit configured to …”.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f), they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f), applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recites sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f).
Claim Rejections - 35 USC § 102
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 –
(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.
Claims 16-30 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Fournarakis.
Fournarakis, Marios, and Markus Nagel. "In-hindsight quantization range estimation for quantized training." Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021. Cited by Applicant in IDS dated 5/7/24.
Regarding claims 16, 24 and 25, Fournarakis discloses a computer-implemented method (and a related device and computer-readable recording medium) of determining a quantization range for tensors of an artificial neural network, the method comprising:
P. 1, sec. 1, “Dynamic quantization can reduce the quantization error as the quantization grid is better utilized but comes with significant memory overhead [4, 8, 20]: the quantization range depends on the full tensor output, therefore, the entire full precision tensor needs to be written to memory before it can be quantized. For typical layers in common DNNs, this can lead up to 8× more memory transfer.”
observing a saturation ratio at a current iteration from the tensors and a quantization range of the artificial neural network; and
P. 4, fig. 3 (reproduced below).
PNG
media_image1.png
466
348
media_image1.png
Greyscale
P. 4, sec. 4, “Figure 3 shows a general framework of how in-hindsight range estimation can be implemented in hardware. The benefit of this approach is that the pre-computed quantization enables fast and efficient static quantization. The required statistics should be easy to calculate at the accumulator or quantization level, to reduce the computational overhead of the method. Such statistics can be the min and max statistics or the saturation ratio 1.” (Emphasis added.)
adjusting the quantization range such that the observed saturation ratio follows a preset target saturation ratio.
The examiner interprets “preset target saturation ratio” according to its broadest reasonable interpretation as encompassing the quantization range specified in Fournarakis as described below.
P. 4, sec. 4, “To quantize the tensor at step t, we use the estimate of quantization ranges from the previous iteration. While the output is computed, appropriate logic keeps track of the min-max statistics from the accumulator. These statistics are then used to update the quantization ranges for the next iteration as soon as the complete output tensor has been calculated. The quantization ranges are calculated as: [see equations below]”.
PNG
media_image2.png
94
458
media_image2.png
Greyscale
Regarding claim 17 and 27, Fournarakis discloses the further limitation wherein the observing of the saturation ratio comprises calculating the ratio of the number of tensors outside the quantization range to the number of tensors.
P. 4, first col., “Such statistics can be the min and max statistics or the saturation ratio 1”.
Footnote 1: “The proportion of values that lie outside the quantization grid.”
Regarding claim 18, Fournarakis discloses the further limitation wherein the adjusting of the quantization range comprises:
calculating a current moving average based on the observed saturation ratio and a past moving average calculated from saturation ratios observed at previous iterations; and
P. 4, sec. 4.1, “We propose an instance of our framework that uses the min-max statistics, which we call in-hindsight min-max. In this method, we define the quantization range as the exponential moving average of the tensor’s min-max statistics.”
See also eqns. 2-3 (reproduced supra).
adjusting the quantization range based on a difference between the current moving average and the target saturation ratio.
Id.
P. 2, first col., “We use a moving average of the quantization range and update the ranges with statistics extracted from the accumulator in an online fashion.”
P. 4, sec. 4.1. “InHindsight MinMax We propose an instance of our framework that uses the min-max statistics, which we call in-hindsight min-max. In this method, we define the quantization range as the exponential moving average of the tensor’s min-max statistics. To quantize the tensor at step t, we use the estimate of quantization ranges from the previous iteration.”
Regarding claim 19, Fournarakis discloses the further limitation wherein
the calculating of the current moving average comprises calculating the current moving average through a weighted sum of the past moving average and the observed saturation ratio.
P. 4, eqns. 2 and 3 (reproduced supra).
The examiner notes that η functions as a weighting term.
Regarding claim 20, Fournarakis discloses the further limitation comprising adjusting a weight of the past moving average and a weight of the observed saturation ratio.
P. 4, eqns. 2 and 3 (reproduced supra).
The examiner notes that η functions as a weighting term.
Regarding claim 21, Fournarakis discloses the further limitation wherein the adjusting of the quantization range comprises:
calculating an amount of change in the quantization range based on the difference between the current moving average and the target saturation ratio; and
P. 5, first col., “We also experimented with using an exponential moving average of the gradient variance [15] to define the quantization ranges.”
adjusting the quantization range according to the amount of change in the quantization range.
P. 4, eqns. 2 and 3 (reproduced supra).
Regarding claim 22, Fournarakis discloses the further limitation comprising setting an initial value of the quantization range based on batch normalization parameters of the artificial neural network.
P. 2, sec. 2.1, ‘BatchNorm’.
Regarding claim 23, Fournarakis discloses the further limitation wherein the tensors are derived from either training data in a training stage of the artificial neural network or user data in an inference stage.
P. 4, sec. 4, “Our proposed method aims at preventing the need for dynamic quantization during quantized training”. (Emphasis added.)
Regarding claims 26, 29 and 30, Fournarakis discloses a computer-implemented method (and a related device) comprising:
receiving information on a quantization range from the outside; and
P. 4, fig. 3 (reproduced supra).
P. 4, sec. 4, “Figure 3 shows a general framework of how in-hindsight range estimation can be implemented in hardware. The benefit of this approach is that the pre-computed quantization enables fast and efficient static quantization. The required statistics should be easy to calculate at the accumulator or quantization level, to reduce the computational overhead of the method. Such statistics can be the min and max statistics or the saturation ratio 1.” (Emphasis added.)
P. 4, sec. 4, “To quantize the tensor at step t, we use the estimate of quantization ranges from the previous iteration. While the output is computed, appropriate logic keeps track of the min-max statistics from the accumulator. These statistics are then used to update the quantization ranges for the next iteration as soon as the complete output tensor has been calculated. The quantization ranges are calculated as: [see equations below]”.
PNG
media_image2.png
94
458
media_image2.png
Greyscale
quantizing tensors of an artificial neural network based on the information on the quantization range, wherein the quantization range is adjusted such that a observed saturation ratio from the quantized tensors of the artificial neural network at a current iteration follows a preset target saturation ratio.
Id.
Regarding claim 28, Fournarakis discloses the further limitation wherein the quantization range is adjusted based on a difference between a current moving average and the target saturation ratio at the current iteration, wherein the current moving average is calculated based on and the observed saturation ratio and a past moving average calculated from saturation ratios observed at previous iterations.
P. 4, sec. 4, “To quantize the tensor at step t, we use the estimate of quantization ranges from the previous iteration. While the output is computed, appropriate logic keeps track of the min-max statistics from the accumulator. These statistics are then used to update the quantization ranges for the next iteration as soon as the complete output tensor has been calculated. The quantization ranges are calculated as: [see equations below]”.
See also p. 4, eqns. 2 and 3 (reproduced supra).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Vincent Gonzales whose telephone number is (571) 270-3837. The examiner can normally be reached on Monday-Friday 7 a.m. to 4 p.m. MT. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Miranda Huang, can be reached at (571) 270-7092.
Information regarding the status of an application may be obtained from the USPTO 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.
/Vincent Gonzales/Primary Examiner, Art Unit 2124
Prosecution Insights