DETAILED ACTION
This action is responsive to the claims filed on 06/13/2024. Claims 1-20 are pending for examination.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 03/09/2026 and 06/13/2024 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-4, 7-11, and 14-18 are rejected under 35 U.S.C. 103 as being unpatentable over Wocjan et al., (Berg, E. V. D., & Wocjan, P. (2023). Techniques for learning sparse Pauli-Lindblad noise models. arXiv preprint arXiv:2311.15408.), hereafter referred to as Wocjan, in view of Sarovar et al., (Sarovar, M., Proctor, T., Rudinger, K., Young, K., Nielsen, E., & Blume-Kohout, R. (2020). Detecting crosstalk errors in quantum information processors. Quantum, 4, 321.), hereafter referred to as Sarovar,
Claim 1: Wocjan teaches the following:
A method for learning noise models to perform quantum error mitigation, (Wocjan, abstract, “Error-mitigation techniques such as probabilistic error cancellation and zero-noise extrapolation benefit from accurate noise models. The sparse Pauli-Lindblad noise model is one of the most successful models for those applications. In existing implementations, the model decomposes into a series of simple Pauli channels with one- and two-local terms that follow the qubit topology.… For such extended models to remain practical, however, we need to ensure that they can be learned efficiently. In this work we present new techniques that accomplish exactly this.… Taken together, these techniques ensure that the learning of the extended noise models remains efficient, despite their increased complexity.”; Page 1, section 1, “A practical way of reducing the effect of noise is the use of error-mitigation techniques.… [T]he second group mostly considers noise associated with the application of gates, and contains techniques such as zero-noise extrapolation … and probabilistic error cancellation.… This technique is leveraged … to mitigate noise associated with Hermitian Clifford operators by learning a sparse Pauli-Lindblad noise model … and subsequently applying the inverse noise map in a quasi-probabilistic manner to obtain unbiased estimates of the expectation value of desired observables.”, Wocjan expressly teaches learning sparse Pauli-Lindblad noise models for use in quantum error-mitigation techniques, including probabilistic error cancellation and zero-noise extrapolation. In particular, Wocjan explains that probabilistic error cancellation mitigates gate-associated noise by first learning a sparse Pauli-Lindblad noise model and subsequently applying the inverse noise map to obtain mitigated estimates of desired observables. Accordingly, Wocjan teaches learning noise models to perform quantum error mitigation, as recited.)
and learning said noise models for each of said set of sub-layers based on said reduced set of learning layers. (Wocjan, abstract, “In existing implementations, the model decomposes into a series of simple Pauli channels with one- and two-local terms that follow the qubit topology.… We introduce twirling based on Pauli rotations, which enables us to automatically generate single-qubit learning correction sequences and reduce the number of unique fidelities that need to be learned. In addition, we propose a basis-selection strategy that leverages graph coloring and uniform covering arrays to minimize the number of learning bases. Taken together, these techniques ensure that the learning of the extended noise models remains efficient, despite their increased complexity.”; page 2, section 2, paragraph 1, “By restricting K to the set of all one- and two-local Pauli terms following the qubit topology, we obtain a sparse Pauli-Lindblad noise model.… Learning the noise model parameters based on a vector of estimated fidelities \hat f can therefore be done using nonnegative least-squares optimization.”, Wocjan decomposes a larger noise model into one- and two-local model components associated with portions of the qubit topology. Wocjan then reduces the number of learning configurations by minimizing the number of learning bases and uses the fidelities obtained through those reduced learning configurations to solve for the individual component-model parameters. Accordingly, Wocjan teaches learning the respective sub-layer noise-model components based upon a reduced collection of learning configurations.)
Sarovar, in the same field of quantum information processing, teaches the following which Wocjan fails to teach:
the method comprising: dividing each target layer of a quantum circuit into a set of sub-layers; (Sarovar, page 4, figure 1, “The dynamics are Markovian and crosstalk-free if the gate operations are modular: the CPTP map describing a given circuit layer can be written as a tensor product of CPTP maps describing each of the component gates (locality), and these com ponent maps do not depend on the other gates in the layer (independence).”; Page 4, section 4.1.2, “Circuits have a well-defined notion of time, which usually defines a natural division into consecutive layers of parallel operations … Operations within a single layer are effectively simultaneous. A layer is uniquely defined by the list of operations applied to each qubit during that layer.”; Page 5, section 4.1.3, “Layer operations can reliably be composed by combining even smaller operations, that act locally and independently. Earlier, we said that a processor is crosstalk-free if it obeys locality and independence. If it is also Markovian, then the conditions for locality and independence can be stated as explicit conditions on the CPTP maps de scribing circuit layers. Each ideal circuit layer defines a locality (tensor product) structure that partitions the qubits into dis joint and uncoupled target subsets.”, Sarovar teaches that a quantum-circuit layer is composed of multiple component gate operations and that the map describing the layer may be decomposed into separate local maps corresponding to those component gates. Sarovar further describes the layer as partitioning the qubits into disjoint target subsets. Under a broad but reasonable interpretation, decomposing the target-layer map into the local maps of its component gates teaches dividing each target layer into a set of sub-layers corresponding to the smaller gate components.)
grouping each of said set of sub-layers for each target layer of said quantum circuit into a reduced set of learning layers; (Sarovar, page 14, section 6.3, “Given a partition of a QIP into regions, we must define a set of circuits to run on the QIP that constitute the crosstalk detection experiment. We only consider circuits that do not (intentionally) couple regions, which means that for each region there is a well-defined subcircuit comprising all operations applied to it… Each possible circuit on the QIP is composed of the parallel application of multiple subcircuits, one on each region.”; Page 14, section 6.3, “Unfortunately, this experiment defines a hypercube containing NM circ distinct circuits, which grows too rapidly with M (the number of regions) to be feasible. However, we observe that in the exhaustive experiment, each subcircuit on every region ri is performed in exponentially many distinct contexts (defined by the settings on the other regions rj ri). This is arduous and overkill; since crosstalk errors are not likely to only be present in one or few of this exponential number of contexts (this is discussed further in Sec. 6.5), we can subsample from this exhaustive experiment. So we will choose a sparse subset of the experiments in the hypercube defining the exhaustive experiment, with the goal of defining a small set of experiments that allow low-weight crosstalk errors to be detected.”, Sarovar teaches defining individual subcircuits for respective disjoint regions and then forming each experimental circuit through the parallel combination of those subcircuits. Rather than executing every possible combination, Sarovar selects a sparse subset that preserves the required characterization information. Thus, the individual subcircuits are grouped into a reduced set of parallel experimental configurations.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the quantum noise model learning techniques of Wocjan with the circuit decomposition and crosstalk-aware characterization techniques taught by Sarovar. Wocjan teaches learning sparse noise models for quantum gate layers to enable scalable probabilistic error cancellation and recognizes the importance of accurately characterizing correlated noise and crosstalk occurring during execution of quantum circuits. However, Wocjan does not specifically describe organizing component gate operations into reduced experimental groupings based upon crosstalk relationships. Sarovar teaches decomposing quantum circuit layers into localized component operations, partitioning gate operations into independent regions, and constructing reduced characterization experiments that preserve crosstalk information while significantly reducing the total number of experiments required. One of ordinary skill in the art would have recognized that incorporating Sarovar’s crosstalk-aware decomposition and reduced characterization methodology into Wocjan’s scalable noise-learning framework would improve characterization efficiency, reduce experimental overhead, and allow accurate learning of layer noise models while preserving correlated error behavior. Such a modification merely applies a known experimental optimization technique to Wocjan’s known noise-model learning process to obtain the predictable result of learning equivalent noise information using fewer characterization experiments.
Claim 2: Wocjan and Sarovar, teaches the limitations of claim 1, Sarovar further teaches:
The method as recited in claim 1, wherein each sub-layer in said set of sub-layers comprises each single and two-qubit gate in a target layer of said quantum circuit. (Sarovar, page 13, col. 1, paragraph 2. “Crosstalk errors are associated with individual layers of elementary operations. In almost every QIP architecture, each elementary operation targets only 1 or 2 qubits. So, since we focus on low-weight crosstalk errors, it is sufficient to consider partitions into disjoint one- and two-qubit regions… We say that a region is allowed if it is possible to define circuits that couple all the qubits within that region, without involving any other qubits. So a 2-region is allowed only if the QIP has a 2-qubit gate directly between those qubits.”, Sarovar teaches decomposing circuit-layer operations according to disjoint one- and two-qubit regions because the elementary operations of the quantum processor are single- and two-qubit operations. A one-qubit region corresponds to a single-qubit gate component, and an allowed two-qubit region corresponds to a two-qubit gate acting directly between the two qubits. Sarovar therefore teaches forming the component sub-layers from the single- and two-qubit gates appearing in the target layer.)
Claim 3: Wocjan and Sarovar, teaches the limitations of claim 1, Wocjan further teaches:
The method as recited in claim 1 further comprising: grouping each of said set of sub-layers for each target layer of said quantum circuit using a gate crosstalk graph. (Wocjan, page 10, paragraph 3, “In order to use these covering arrays, we start the basis-selection protocol by generating a graph with vertices corresponding to the qubits in the noise model and edges for the neighboring qubits, which can include virtual connections that indicate expected crosstalk terms between the given pairs of qubits. Once the graph has been generated we apply graph coloring and find a covering array.… By mapping each of the colors to a column in the covering array we can then construct N bases.”
Page 11, paragraph 1, “The implication of these simplifications is that we can merge any pair of vertices in the graph whose qubits are subject to the same two-qubit gate.… Following these transformations we can apply graph coloring and determine the bases using a binary covering array.”,
Page 12, figure 4, “Application of the learning protocol to … a layer of gates CZ(1,2), CX(3,4), CZ(5,6), and Swap(7,8) on an example topology with presumed crosstalk between qubits 2 and 5.… We start by finding a graph coloring for the qubit topology.… [W]e first merge the vertices of the qubits of each gate … [and] color the resulting graph … to obtain the different bases.”, Wocjan expressly constructs a graph in which vertices correspond to the physical qubits participating in gate operations and edges represent topology relationships, including additional virtual edges for expected crosstalk. Wocjan then merges vertices belonging to the same two-qubit gate and uses graph coloring to arrange the resulting gate-associated components into compatible learning configurations. The resulting graph is therefore a gate crosstalk graph under the broadest reasonable interpretation, and the graph is expressly used to group the gate-associated model components for learning.)
Claim 4: Wocjan and Sarovar, teaches the limitations of claim 1, Sarovar further teaches:
The method as recited in claim 1, wherein said grouping of each of said set of sub-layers for each target layer of said quantum circuit into said reduced set of learning layers comprises combining all parallelizable components in each of said set of sub-layers for each target layer of said quantum circuit. (Sarovar, page 13, col. 1, paragraph 2. “Crosstalk errors are associated with individual layers of elementary operations. In almost every QIP architecture, each elementary operation targets only 1 or 2 qubits. So, since we focus on low-weight crosstalk errors, it is sufficient to consider partitions into disjoint one- and two-qubit regions… We say that a region is allowed if it is possible to define circuits that couple all the qubits within that region, without involving any other qubits. So a 2-region is allowed only if the QIP has a 2-qubit gate directly between those qubits.”, Sarovar teaches decomposing circuit-layer operations according to disjoint one- and two-qubit regions because the elementary operations of the quantum processor are single- and two-qubit operations. A one-qubit region corresponds to a single-qubit gate component, and an allowed two-qubit region corresponds to a two-qubit gate acting directly between the two qubits. Sarovar therefore teaches forming the component sub-layers from the single- and two-qubit gates appearing in the target layer.)
Claim 7: Wocjan and Sarovar, teaches the limitations of claim 1, Sarovar further teaches:
The method as recited in claim 1, wherein said quantum circuit is unstructured. (Sarovar, page 3, col. 2, paragraph 1, “Definition 1: A QIP’s behavior is crosstalk-free if its behavior, when implementing arbitrary circuits, satisfies locality and independence.”; Page 4, section 4.1.2, paragraph 1-2, “We need a stronger notion of modularity to predict how a QIP will perform on new quantum circuits that have not been run before. Circuits have a well-defined notion of time, which usually defines a natural division into consecutive layers 1 of parallel operations (gates, state preparations or measurements). See Fig. 1 for an example circuit with 9 layers that we notate L0,..., L8. Operations within a single layer are effectively simultaneous. A layer is uniquely defined by the list of operations applied to each qubit during that layer, where “operations” can include idles, measurements, and initialiation/reset operations as well as elementary gates. Figure 1(b) shows a circuit partitioned into layers. We call the QIP Markovian if we can describe and model each unique layer by a CPTP map acting on all n qubits in the system.”, Sarovar applies its characterization method to arbitrary and previously unexecuted quantum circuits and expressly recognizes that such circuits may contain exponentially many possible or unique layers. Because the present application defines an unstructured circuit as one containing many unique layers, Sarovar teaches or at least renders obvious applying the combined characterization method to an unstructured quantum circuit.)
Claims 8-11 and 14 recite limitations substantially similar to claims 1-4 and 7, as such a similar analysis applies.
Claim 8 recites the following additional limitation for consideration which Wocjan further teaches:
A computer program product for learning noise models to perform quantum error mitigation, the computer program product comprising one or more computer readable storage mediums having program code embodied therewith, the program code comprising programming instructions for: (Wocjan, abstract, “Error-mitigation techniques such as probabilistic error cancellation and zero-noise extrapolation benefit from accurate noise models. The sparse Pauli-Lindblad noise model is one of the most successful models for those applications.… For such extended models to remain practical, however, we need to ensure that they can be learned efficiently. In this work we present new techniques that accomplish exactly this.… We introduce twirling based on Pauli rotations, which enables us to automatically generate single-qubit learning correction sequences and reduce the number of unique fidelities that need to be learned. In addition, we propose a basis-selection strategy that leverages graph coloring and uniform covering arrays to minimize the number of learning bases.”; Page 2, section 2, paragraph 2, “Learning the noise model parameters based on a vector of estimated fidelities f can therefore be done using nonnegative least-squares optimization:
PNG
media_image1.png
44
460
media_image1.png
Greyscale
”; page 3, paragraph 2, “In Section 5 we present an efficient basis-selection algorithm that applies for our extended noise models and includes the existing learning algorithm as a special case.”, Wocjan teaches an algorithmic procedure for learning quantum noise models for use in quantum error mitigation, including automatically generating learning correction sequences, performing nonnegative least-squares optimization to determine noise-model parameters, and executing an efficient basis-selection algorithm. Although Wocjan does not expressly identify the conventional computer-readable storage medium on which the instructions implementing these algorithms are stored, one of ordinary skill in the art would have understood that Wocjan’s expressly disclosed automatic algorithms and numerical optimization procedures may predictably be implemented as program instructions stored on a computer-readable storage medium for execution by a computer. Implementing Wocjan’s disclosed computational noise-model-learning algorithms as stored program code would have amounted to use of a known computer implementation technique according to its established function and would not have altered the operation of Wocjan’s noise-learning method. Accordingly, Wocjan teaches or renders obvious a computer program product comprising one or more computer-readable storage mediums having program code embodied therewith for performing the recited noise-model-learning operations.)
Claims 15-18 recite limitations substantially similar to claims 1-4, as such a similar analysis applies.
Claim 15 recites the following additional limitation for consideration which Wocjan further teaches:
A system, comprising: a memory for storing a computer program for learning noise models to perform quantum error mitigation; and a processor connected to said memory, wherein said processor is configured to execute program instructions of the computer program comprising: (Wocjan, abstract, “Error-mitigation techniques such as probabilistic error cancellation and zero-noise extrapolation benefit from accurate noise models. The sparse Pauli-Lindblad noise model is one of the most successful models for those applications.… For such extended models to remain practical, however, we need to ensure that they can be learned efficiently. In this work we present new techniques that accomplish exactly this.… We introduce twirling based on Pauli rotations, which enables us to automatically generate single-qubit learning correction sequences and reduce the number of unique fidelities that need to be learned. In addition, we propose a basis-selection strategy that leverages graph coloring and uniform covering arrays to minimize the number of learning bases.”; Page 2, section 2, paragraph 2, “Learning the noise model parameters based on a vector of estimated fidelities f can therefore be done using nonnegative least-squares optimization:
PNG
media_image1.png
44
460
media_image1.png
Greyscale
”; page 3, paragraph 2, “In Section 5 we present an efficient basis-selection algorithm that applies for our extended noise models and includes the existing learning algorithm as a special case.”; page 11, paragraph 1, “Following these transformations we can apply graph coloring and determine the bases using a (binary) covering array.… When extracting the learning bases from the graph coloring and the binary covering array, we expand each symbol for merged vertices to two gate-dependent Pauli terms, one for each qubit associated with the original vertices.”, Wocjan teaches computational algorithms for learning quantum noise models, including automatically generating learning sequences, solving a nonnegative least-squares optimization problem to determine noise-model parameters, applying graph coloring, and algorithmically determining learning bases from covering arrays. Although Wocjan does not expressly recite a processor and memory architecture, one of ordinary skill in the art would have understood that execution of the disclosed automatic algorithms and numerical optimization procedures on a conventional computing system would predictably employ a processor executing stored program instructions and a memory storing those instructions and associated data. Providing a processor and memory for executing Wocjan’s disclosed algorithms would merely constitute implementation of the disclosed computational procedure using conventional computer components according to their established functions, yielding the predictable result of automated execution of Wocjan’s noise-model-learning method. Accordingly, Wocjan teaches or renders obvious a system having a memory storing the computer program and a processor connected to the memory and configured to execute program instructions comprising the recited noise-model-learning operations.)
Claims 5-6, 12-13 and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Wocjan in view of Sarovar, and in further view of Berg et al., (van den Berg, E., Minev, Z. K., Kandala, A., & Temme, K. (2023). Probabilistic error cancellation with sparse Pauli–Lindblad models on noisy quantum processors. Nature Physics, 19(8), 1116–1121. https://doi.org/10.1038/s41567-023-02042-2), hereafter referred to as Berg.
Claim 5: Wocjan and Sarovar, teaches the limitations of claim 1, Berg, in the same field of neural networks, teaches the following which Wocjan and Sarovar fail to teach:
The method as recited in claim 1 further comprising: combining said learned noise models forming a complete set of noise models for target layers of said quantum circuit. (Berg, page 2, col. 1, paragraph 1, “
PNG
media_image2.png
45
329
media_image2.png
Greyscale
where wk = 2−1(1 + e−2λk).The model terms K are chosen to reflect the noise interactions in the quantum processor and their number, which determines the model complexity and expressivity, typically scales polynomially in n and therefore allows us to represent noise models for the full device by a small set of nonnegative coefficients λk.”; Page 5, col. 1, paragraph 1, “The circuit contains two unique layers of cx gates, one starting at even and one at odd locations in the qubit chain. Once the noise models for the two layers are learned, we generate random circuit instances. We apply readout-error mitigation on all observables (see [33] for more on readout mitigation). To counter time-dependent fluctuations in the noise we relearn the noise model after fixed intervals (see also Supplementary Materials Sec. SVII). The final observables are obtained after averaging… All other models were learned in a similar fashion.”, Berg represents the complete noise channel as a product of the individual simple Pauli channels associated with the learned local model terms. The individually learned model components are therefore mathematically combined to form the complete noise model for a layer. Berg also learns respective models for each unique gate layer appearing in the target circuit. The resulting collection of reconstructed layer models constitutes a complete set of noise models for the target layers of the circuit.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to further modify the combined teachings of Wocjan and Sarovar with the teachings of Berg. Wocjan and Sarovar collectively teach efficient learning and characterization of quantum noise models through decomposition of quantum operations, reduced learning configurations, and crosstalk-aware grouping of component operations. Berg further teaches that learned noise-model components may be combined to represent the noise associated with target layers of a quantum circuit and that, once such noise models have been learned, the models may be used to mitigate noise through probabilistic error cancellation by implementing corresponding inverse noise channels. One of ordinary skill in the art would have been motivated to incorporate Berg’s techniques into the noise-model-learning framework of Wocjan and Sarovar in order to obtain complete noise models corresponding to the target circuit layers and utilize the learned models for their known and intended purpose of mitigating errors during quantum-circuit execution. Such a combination would have predictably provided an efficient characterization framework in which reduced, crosstalk-aware learning configurations are used to learn the constituent noise information, the learned information is combined to obtain the applicable target-layer noise models, and those models are subsequently used to perform quantum error mitigation.
Claim 6: Wocjan and Sarovar, teaches the limitations of claim 1, Berg, in the same field of neural networks, teaches the following which Wocjan and Sarovar fail to teach:
The method as recited in claim 1 further comprising: performing quantum error mitigation on said quantum circuit using said learned noise models. (Berg, page 3, col. 2, last paragraph, “Probabilistic error cancellation Once the noise model has been learned, it can be used to mitigate the noise using the PEC method[9].The protocol implements the channelinverseΛ−1 i through quasi-probabilistic sampling for each of the l layers. The inverse of the map Λ is obtained by negating L, leading to a non-physical map”; Supplementary Information, section SI., “Mitigation[:] •Given a circuit that contains the layer of gates •Generate multiple circuit instances with each layer preceded by a Pauli sampled from the quasi-probability distribution and with a Pauli twirled instance of the layer •Estimate the expectation of the observables of interest and scale by γ”, Berg expressly uses the learned layer-specific noise models to determine the inverse noise channels and implements those inverses through quasi-probabilistic sampling. That operation is probabilistic error cancellation and therefore directly teaches performing quantum error mitigation on the target quantum circuit using the learned noise models.)
The rationale to combine Berg with Wocjan and Sarovar is similar to that as applied for claim 5 above.
Claims 12-13 and 18-20 recite limitations substantially similar to claims 5-6, as such a similar analysis applies.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Nation, P. D., Kang, H., Sundaresan, N., & Gambetta, J. M. (2021). Scalable mitigation of measurement errors on quantum computers. PRX Quantum, 2(4), 040326.
Endo, S., Benjamin, S. C., & Li, Y. (2018). Practical quantum error mitigation for near-future applications. Physical Review X, 8(3), 031027.
Murali, P., McKay, D. C., Martonosi, M., & Javadi-Abhari, A. (2020, March). Software mitigation of crosstalk on noisy intermediate-scale quantum computers. In Proceedings of the twenty-fifth international conference on architectural support for programming languages and operating systems (pp. 1001-1016).
Li, G., Ding, Y., & Xie, Y. (2019, April). Tackling the qubit mapping problem for NISQ-era quantum devices. In Proceedings of the twenty-fourth international conference on architectural support for programming languages and operating systems (pp. 1001-1014).
Van, Minev, Z. K., Abhinav Kandala, & Temme, K. (2022). Probabilistic error cancellation with sparse Pauli-Lindblad models on noisy quantum processors. arXiv (Cornell University). doi.org/10.48550/arxiv.2201.09866
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HYUNGJUN B YI whose telephone number is (703)756-4799. The examiner can normally be reached M-F 9-5.
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, Usmaan Saeed can be reached on (571) 272-4046. 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.
/H.B.Y./Examiner, Art Unit 2124
/USMAAN SAEED/Supervisory Patent Examiner, Art Unit 2146
Prosecution Insights