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 . Claims 1-10 are presented in the case.
Priority
Acknowledgment is made of applicant's claim for foreign priority based on PCT application CN202310391219.8 filed in China on 04/12/2023. Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Information Disclosure Statement
The information disclosure statement submitted on 04/08/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Objections
Claims 2, 5 and 7, 10 are objected to because of the following informalities:
Claim 2, line 3 recites the phrase “behind the floating-point layer” which should be “behind after a respective one of the floating-point layers”
Claim 5, line 2 recites the phrase “wherein data format of the floating-point layer is: floating-point FP32 or floating-point FP16.” which should be “wherein each of the floating-point layers has a data format selected from FP32 and FP16”
Claim 7, line 3 recites the phrase “behind the floating-point layer” which should be “behind after a respective one of the floating-point layers”
Claim 10, line 2 recites the phrase “wherein data format of the floating-point layer is: floating-point FP32 or floating-point FP16.” which should be “wherein each of the floating-point layers has a data format selected from FP32 and FP16”
For the informalities above and wherever else they may occur appropriate correction is required.
Claim Interpretation
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.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
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.
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) or pre-AIA 35 U.S.C. 112, sixth paragraph, 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: an acquisition module, configured to obtain a floating-point model with multiple floating-point layers, and calculate a cumulative original output of all floating-point layers of the floating-point model;
a quantization module, configured to select one floating-point layer from the floating-point model separately each time for quantization, so as to form multiple hybrid models each containing one quantization layer, and separately calculate an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error value;
a sorting module, configured to sort the calculated error values; and
a restoring module, configured to quantize all floating-point layers of the floating-point model, and restore corresponding quantization layer(s) to floating-point layer(s) one by one in descending order of the calculated error values and calculate a difference between cumulative output of all layers of a corresponding restored model and the cumulative original output until the difference is less than a preset loss threshold to obtain a target hybrid model in claims 6-10.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, 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 this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (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 it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
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 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”)
Claims 1 and 6 have the following abstract idea analysis.
Step 1: The claims are directed to “a method and system. The claims are directed to the statutory categories accordingly.
Step 2A Prong 1: claim recites the abstract idea limitations of "calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values" and "quantizing all floating-point layers". The limitations include a mathematical concept see MPEP § 2106.04(a)(2)) where it cites textual recitation can still be mathematical "determining a ration of A to B". The specification also provides example math using Fourier and cosine functions (See USPGPUB ¶45-46). See USPTO 2024 example 48 where STFT conversion and determining vectors by formula were treated as mathematical operations. Thus, the limitations are an abstract idea in the “mathematical concept”. Other sections of the claims such as "models" "acquisition module" and other "modules" are advanced processes, too generic or high level to be listed as a judicial exception given the available descriptions and MPEP comparisons.
Step 2A Prong 2: The judicial exceptions recited in these claims are not integrated into a practical application. Merely invoking "models" "acquisition module" and other "modules" does not yield eligibility. Claims are still in line with mathematical concepts such as claims 1 and 6 are not specific to a practical application. The additional elements as such are processors and instructions which do not include specialized hardware. See MPEP § 2106.05(a). The math is just being used to produce a result. Claims 1 and 6 do not include a more specific field but even doing so may not be sufficient to overcome the abstract idea rejection. Merely applying an math to a field without an advancement in the new field or new hardware is ineligible. See MPEP § 2106.05(h).
Step 2B: The claims do not contain significantly more than their judicial exceptions. Processors, memory and other hardware are in their standard forms in the field. Note generic processors are recited not new quantum processors. These additional elements are well-understood, routine, and conventional activity, see MPEP 2106.05(d)(II). Claims lacks any particular "how" or algorithm for a solution in a field in a novel way. Claims require more specificity on processes that would be incapable of simple mathematics, mental processes or use more substantial structure than conventional devices such as non-textbook implementations. Specification USPGPUB ¶4 does state that the invention increases speed of models but the claim must reflect the mechanism responsible for the improvement.
Regarding claims 2-5 and 6-10 they merely narrow the previously recited abstract idea limitations with more abstract concepts and/or routine fundamental processes. For the reasons described above with respect to claim 1 and 6 this judicial exception is not meaningfully integrated into a practical application, or significantly more than the abstract idea. Abstract idea steps 1, 2A prong 1 and 2 remain the same as independent analysis above. See specification for more practical application concepts as none are seen in claims 2-5 and 6-10.
With respect to step 2B These claims disclose similar limitations described for the dependent claims above and do not provide anything significantly more than organizing human activity concepts. Claims 2-5 and 6-10 recite the additional elements of "wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layer and used to normalize an output of the floating-point layer. wherein before quantizing all floating-point layers of the floating-point model, the restoring module is further configured to remove the last layer that is a normalized exponential function layer of the floating-point model. wherein upon the difference being less than the preset loss threshold, the restoring module is further configured to add a normalized exponential function layer to the target hybrid model as the last layer of the target hybrid model. wherein data format of the floating-point layer is: floating-point FP32 or floating-point FP16.". These elements are more abstract concepts, generic applications to a field of use or well-understood, routine, conventional activity (see MPEP § 2106.05(d) and can't be simply appended to qualify as significantly more or being a practical application. What type of application, or structure of components beyond generic machine learning is still unknown for these claims. Therefore claims 2-5 and 6-10 also recites abstract ideas that do not integrate into a practical application or amount to significantly more than the judicial exception, and are rejected under U.S.C. 101.
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 of this title, 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, 5-6 and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Qadeer et al. (US 20210174172 A1) hereinafter Qadeer in view of Shen et al. (US 20220129736 A1) hereinafter Shen.
As to independent claim 1, Qadeer teaches a method for improving accuracy of model quantization, comprising: [neural network quantizing ¶6]
obtaining a floating-point model with multiple floating-point layers, and [floating point network (model) with layers ¶6, ¶10 "floating-point networks represented as a set of layers"]
[[quantization]]
[selectively converts layer bit width via quantization ¶12-13 " convert each set of floating-point weights for each floating-point layer of the floating-point network in a set of low-bit-width weights. Therefore, the quantized network initially includes a set of low-bit-width layers"] so as to form multiple hybrid models each containing one quantization layer, and [hybrid quantized network with low-bit (quant layer) ¶9 "system generates a hybrid quantized network characterized by a select set of layers represented at a high bit-width (e.g., sixteen-bit fixed-point), while most layers of the network are represented at a low-bit-width (e.g., eight-bit fixed-point)"]
sorting the calculated error values; and [sorts, orders or ranks according to accuracy impact and contribution on accuracy or deviation (impacting accuracy impacts error) ¶41, ¶46 ¶50 " sort the low-bit-width layers of the low-bit-width quantized network based on the likelihood that these layers are reducing the overall accuracy of the quantized network by measuring their impact on the full network accuracy"]
quantizing all floating-point layers of the floating-point model, and [quants each layer ¶12 "generates this fully quantized network"]
restoring corresponding quantization layer(s) to floating-point layer(s) one by one in descending order of the calculated error values and [iterates through low bit layers and replaces with high (restores) based on deviation ¶12-14 "system can iterate through these ordered low-bit-width layers starting with the low-bit-width layer with the greatest per-layer deviation from its corresponding floating-point layer"]
calculating a difference between cumulative output of all layers of a corresponding restored model and the cumulative original output until the difference is less than a preset loss threshold to obtain a target hybrid model. [reevaluates and iterates until the network satisfies the threshold ¶14 "system continues this iterative process until the accuracy of this “hybrid” quantized network (i.e. a network containing layers quantized at low-bit-width and high bit-width) satisfies the loss-of-accuracy threshold provided by the user"]
Qadeer does not specifically teach calculating a cumulative original output of all floating-point layers of the floating-point model; selecting one floating-point layer from the floating-point model separately each time for quantization, separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values.
However, Shen teaches calculating a cumulative original output of all floating-point layers of the floating-point model; [sequential L1 layer to L3 layer are added to a original final output that uses floating point precision layers, Fig. 1 ¶6, ¶16 "X4 is the final output of the neural network and is referred as the original final output"]
selecting one floating-point layer from the floating-point model separately each time for quantization, [quantization on one layer at a time (L1 selected first) Fig. 4-6, ¶18, ¶6 "For a particular layer of the plurality of layer, quantization of a second precision on the particular layer and an input of the particular layer is performed"]
separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values; [calculates SQNR, cross entropy, and divergence (error) of final and original final output ¶23 "obtains the value of the objective function LS1 according to the final output X4′ and the original final output X4. The objective function LS1 can be signal-to-quantization-noise ratio (SQNR), cross entropy, cosine similarity, or KL divergence (Kullback-Leibler divergence). However, the present invention is not limited thereto, and any functions capable of calculating the loss between the final output X4"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the quantization disclosed by Qadeer by incorporating the calculating a cumulative original output of all floating-point layers of the floating-point model; selecting one floating-point layer from the floating-point model separately each time for quantization, separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values by Shen because both techniques address the same field of machine learning and by incorporating Shen into Qadeer better balances the cost and prediction precision for neural networks for improved processes [Shen ¶3-4]
As to dependent claim 5, the rejection of claim 1 is incorporated, Qadeer and Shen further teach wherein data format of (each)the floating-point layer is: floating-point FP32 or floating-point FP16. [Shen FP32 ¶16]
As to independent claim 6, Qadeer teaches a system for improving accuracy of model quantization, comprising: [system and neural network quantizing ¶6, ¶9]
an acquisition module, a quantization module, a sorting module, and a restoring module, [processor that performs quantization, acquisition, sorting and restoring ¶9, ¶66]
obtaining a floating-point model with multiple floating-point layers, and [floating point network (model) with layers ¶6, ¶10 "floating-point networks represented as a set of layers"]
[[quantization]]
[selectively converts layer bit width via quantization ¶12-13 " convert each set of floating-point weights for each floating-point layer of the floating-point network in a set of low-bit-width weights. Therefore, the quantized network initially includes a set of low-bit-width layers"] so as to form multiple hybrid models each containing one quantization layer, and [hybrid quantized network with low-bit (quant layer) ¶9 "system generates a hybrid quantized network characterized by a select set of layers represented at a high bit-width (e.g., sixteen-bit fixed-point), while most layers of the network are represented at a low-bit-width (e.g., eight-bit fixed-point)"]
sorting the calculated error values; and [sorts, orders or ranks according to accuracy impact and contribution on accuracy or deviation (impacting accuracy impacts error) ¶41, ¶46 ¶50 " sort the low-bit-width layers of the low-bit-width quantized network based on the likelihood that these layers are reducing the overall accuracy of the quantized network by measuring their impact on the full network accuracy"]
quantizing all floating-point layers of the floating-point model, and [quants each layer ¶12 "generates this fully quantized network"]
restoring corresponding quantization layer(s) to floating-point layer(s) one by one in descending order of the calculated error values and [iterates through low bit layers and replaces with high (restores) based on deviation ¶12-14 "system can iterate through these ordered low-bit-width layers starting with the low-bit-width layer with the greatest per-layer deviation from its corresponding floating-point layer"]
calculating a difference between cumulative output of all layers of a corresponding restored model and the cumulative original output until the difference is less than a preset loss threshold to obtain a target hybrid model. [reevaluates and iterates until the network satisfies the threshold ¶14 "system continues this iterative process until the accuracy of this “hybrid” quantized network (i.e. a network containing layers quantized at low-bit-width and high bit-width) satisfies the loss-of-accuracy threshold provided by the user"]
Qadeer does not specifically teach calculating a cumulative original output of all floating-point layers of the floating-point model; selecting one floating-point layer from the floating-point model separately each time for quantization, separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values.
However, Shen teaches calculating a cumulative original output of all floating-point layers of the floating-point model; [sequential L1 layer to L3 layer are added to a original final output that uses floating point precision layers, Fig. 1 ¶6, ¶16 "X4 is the final output of the neural network and is referred as the original final output"]
selecting one floating-point layer from the floating-point model separately each time for quantization, [quantization on one layer at a time (L1 selected first) Fig. 4-6, ¶18, ¶6 "For a particular layer of the plurality of layer, quantization of a second precision on the particular layer and an input of the particular layer is performed"]
separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values; [calculates SQNR, cross entropy, and divergence (error) of final and original final output ¶23 "obtains the value of the objective function LS1 according to the final output X4′ and the original final output X4. The objective function LS1 can be signal-to-quantization-noise ratio (SQNR), cross entropy, cosine similarity, or KL divergence (Kullback-Leibler divergence). However, the present invention is not limited thereto, and any functions capable of calculating the loss between the final output X4"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the quantization disclosed by Qadeer by incorporating the calculating a cumulative original output of all floating-point layers of the floating-point model; selecting one floating-point layer from the floating-point model separately each time for quantization, separately calculating an error value of cumulative output of all layers of each hybrid model relative to the cumulative original output, so as to obtain multiple calculated error values by Shen because both techniques address the same field of machine learning and by incorporating Shen into Qadeer better balances the cost and prediction precision for neural networks for improved processes [Shen ¶3-4]
As to dependent claim 10, the rejection of claim 6 is incorporated, Qadeer and Shen further teach wherein data format of (each)the floating-point layer is: floating-point FP32 or floating-point FP16. [Shen FP32 ¶16]
Claims 2 and 7 are rejected under 35 U.S.C. 103 as being unpatentable over Qadeer in view of Shen, as applied in claim 1 and 6 above, and further in view of Sivakumar et al. (US 20200134448 A1) hereinafter Sivakumar.
As to dependent claim 2, Qadeer and Shen teach the method of claim 1 above that is incorporated,
Qadeer and Shen do not specifically teach wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer.
However, Sivakumar teaches wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer. [batch normalization layers generate normalized weights of float weights ¶32-34 " batch normalized neural network layer and generates batch normalized weights from the floating point weights using the correction factor"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the neural networks disclosed by Qadeer and Shen by incorporating the wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer disclosed by Sivakumar because all techniques address the same field of machine learning and by incorporating Sivakumar into Qadeer and Shen helps system reduce consumption of resources and retain high performance [Sivakumar ¶6].
As to dependent claim 7, Qadeer and Shen teach the method of claim 6 above that is incorporated,
Qadeer and Shen do not specifically teach wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer.
However, Sivakumar teaches wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer. [batch normalization layers generate normalized weights of float weights ¶32-34 " batch normalized neural network layer and generates batch normalized weights from the floating point weights using the correction factor"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the neural networks disclosed by Qadeer and Shen by incorporating the wherein the floating-point model further comprises batch normalization layers, each disposed behind the floating-point layers and used to normalize an output of the floating-point layer disclosed by Sivakumar because all techniques address the same field of machine learning and by incorporating Sivakumar into Qadeer and Shen helps system reduce consumption of resources and retain high performance [Sivakumar ¶6].
Claims 3-4 and 8-9 are rejected under 35 U.S.C. 103 as being unpatentable over Qadeer in view of Shen, as applied in claim 1 and 6 above, and further in view of Liu et al. (US 20210334646 A1) hereinafter Liu.
As to dependent claim 3, Qadeer and Shen teach the method of claim 1 above that is incorporated,
Qadeer and Shen do not specifically teach wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model.
However, Liu teaches wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model. [discards (removes) softmax layer (last function layer) ¶69-70 "function with respect to the input x for n.sub.K output classes. Here f is referred to as the logit layer. The softmax layer can be safely discarded in the analysis due to its monotonicity"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the neural networks disclosed by Qadeer and Shen by incorporating the wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model disclosed by Liu because all techniques address the same field of machine learning and by incorporating Liu into Qadeer and Shen reduces generalization errors and provides more efficient processing [Liu ¶3-4].
As to dependent claim 4, the rejection of claim 3 is incorporated, Qadeer, Shen and Liu further teach wherein upon the difference being less than the preset loss threshold, the method further comprises: adding a normalized exponential function layer to the target hybrid model as the last layer of the target hybrid model. [Liu sometimes passes output (adds at end last) through activation function ¶70, ¶77 "a perceptron produces a single output based on several real-valued inputs by forming a linear combination using its input weights (and sometimes passing the output through a nonlinear activation function)"]
As to dependent claim 8, Qadeer and Shen teach the method of claim 6 above that is incorporated,
Qadeer and Shen do not specifically teach wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model.
However, Liu teaches wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model. [discards (removes) softmax layer (last function layer) ¶69-70 "function with respect to the input x for n.sub.K output classes. Here f is referred to as the logit layer. The softmax layer can be safely discarded in the analysis due to its monotonicity"]
Accordingly, it would have been obvious to a person of ordinary skill in the art before the effective filling date of the claimed invention to modify the neural networks disclosed by Qadeer and Shen by incorporating the wherein before quantizing all floating-point layers of the floating-point model, the method further comprises removing the last layer that is a normalized exponential function layer of the floating-point model disclosed by Liu because all techniques address the same field of machine learning and by incorporating Liu into Qadeer and Shen reduces generalization errors and provides more efficient processing [Liu ¶3-4].
As to dependent claim 9, the rejection of claim 8 is incorporated, Qadeer, Shen and Liu further teach wherein upon the difference being less than the preset loss threshold, the method further comprises: adding a normalized exponential function layer to the target hybrid model as the last layer of the target hybrid model. [Liu sometimes passes output (adds at end last) through activation function ¶70, ¶77 "a perceptron produces a single output based on several real-valued inputs by forming a linear combination using its input weights (and sometimes passing the output through a nonlinear activation function)"]
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Applicant is required under 37 C.F.R. § 1.111(c) to consider these references fully when responding to this action.
Young (US 20180174022 A1) teaches softmax layers for exponential measure of neural network output (see ¶3-4)
It is noted that any citation to specific pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. A reference is relevant for all it contains and may be relied upon for all that it would have reasonably suggested to one having ordinary skill in the art. In re Heck, 699 F.2d 1331, 1332-33, 216 U.S.P.Q. 1038, 1039 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006, 1009, 158 U.S.P.Q. 275, 277 (C.C.P.A. 1968)).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Beau Spratt whose telephone number is 571 272 9919. The examiner can normally be reached 8:30am to 5:00pm (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, Jennifer Welch can be reached at 571 272 7212. The fax phone number for the organization where this application or proceeding is assigned is 571 483 7388.
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.
/BEAU D SPRATT/ Primary Examiner, Art Unit 2143
Prosecution Insights