Abstract: A modular autoencoder model is described. The modular autoencoder model comprises input models configured to process one or more inputs to a first level of dimensionality suitable for combination with other inputs; a common model configured to: reduce a dimensionality of combined processed inputs to generate low dimensional data in a latent space; and expand the low dimensional data in the latent space into one or more expanded versions of the one or more inputs suitable for generating one or more different outputs; output models configured to use the one or more expanded versions of the one or more inputs to generate the one or more different outputs, the one or more different outputs being approximations of the one or more inputs; and a prediction model configured to estimate one or more parameters based on the low dimensional data in the latent space.

Abstract: A modular autoencoder model is described. The modular autoencoder model comprises input models configured to process one or more inputs to a first level of dimensionality suitable for combination with other inputs; a common model configured to: reduce a dimensionality of combined processed inputs to generate low dimensional data in a latent space; and expand the low dimensional data in the latent space into one or more expanded versions of the one or more inputs suitable for generating one or more different outputs; output models configured to use the one or more expanded versions of the one or more inputs to generate the one or more different outputs, the one or more different outputs being approximations of the one or more inputs; and a prediction model configured to estimate one or more parameters based on the low dimensional data in the latent space.

Abstract: Methods and systems for determining a mapped intensity metric are described. Determining the mapped intensity metric includes determining an intensity metric for a manufacturing system. The intensity metric is determined based on a reflectivity of a location on a substrate and a manufacturing system characteristic. Determining the mapped intensity metric also includes determining a mapped intensity metric for a reference system. The reference system has a reference system characteristic. The mapped intensity metric is determined based on the intensity metric, the manufacturing system characteristic, and the reference system characteristic, to mimic determination of the intensity metric for the manufacturing system using the reference system. In some embodiments, the reference system is virtual, and the manufacturing system is physical.

Abstract: Methods for calibrating metrology apparatuses and determining a parameter of interest are disclosed. In one arrangement, training data is provided that comprises detected representations of scattered radiation detected by each of plural metrology apparatuses. An encoder encodes each detected representation to provide an encoded representation, and a decoder generates a synthetic detected representation from the respective encoded representation. A classifier estimates from which metrology apparatus originates each encoded representation or each synthetic detected representation. The training data is used to simultaneously perform, in an adversarial relationship relative to each other, a first machine learning process involving the encoder or decoder and a second machine learning process involving the classifier.

Abstract: Parameters of a structure (900) are measured by reconstruction from observed diffracted radiation. The method includes the steps: (a) defining a structure model to represent the structure in a two- or three-dimensional model space; (b) using the structure model to simulate interaction of radiation with the structure; and (c) repeating step (b) while varying parameters of the structure model. The structure model is divided into a series of slices (a-f) along at least a first dimension (Z) of the model space. By the division into slices, a sloping face (904, 906) of at least one sub-structure is approximated by a series of steps (904?, 906?) along at least a second dimension of the model space (X). The number of slices may vary dynamically as the parameters vary. The number of steps approximating said sloping face is maintained constant. Additional cuts (1302, 1304) are introduced, without introducing corresponding steps.

Abstract: A method of determining an estimated intensity of radiation scattered by a target illuminated by a radiation source, has the following steps: obtaining and training (402) a library REFLIB of wavelength-dependent reflectivity as a function of the wavelength, target structural parameters and angle of incidence R(?,?,x,y); determining (408) a wide-band library (W-BLIB) of integrals of wavelength-dependent reflectivity R of the target in a Jones framework over a range of radiation source wavelengths ?; training (TRN) (410) the wide-band library; and determining (412), using the trained wide-band library, an estimated intensity (INT) of radiation scattered by the target illuminated by the radiation source.

Abstract: A method, computer program and associated apparatuses for metrology. The method includes determining a reconstruction recipe describing at least nominal values for use in a reconstruction of a parameterization describing a target. The method includes obtaining first measurement data relating to measurements of a plurality of targets on at least one substrate, the measurement data relating to one or more acquisition settings and performing an optimization by minimizing a cost function which minimizes differences between the first measurement data and simulated measurement data based on a reconstructed parameterization for each of the plurality of targets. A constraint on the cost function is imposed based on a hierarchical prior. Also disclosed is a hybrid model method comprising obtaining a coarse model operable to provide simulated coarse data; and training a data driven model to correct the simulated coarse data so as to determine simulated data for use in reconstruction.

Abstract: Parameters of a structure (900) are measured by reconstruction from observed diffracted radiation. The method includes the steps: (a) defining a structure model to represent the structure in a two- or three-dimensional model space; (b) using the structure model to simulate interaction of radiation with the structure; and (c) repeating step (b) while varying parameters of the structure model. The structure model is divided into a series of slices (a-f) along at least a first dimension (Z) of the model space. By the division into slices, a sloping face (904, 906) of at least one sub-structure is approximated by a series of steps (904?, 906?) along at least a second dimension of the model space (X). The number of slices may vary dynamically as the parameters vary. The number of steps approximating said sloping face is maintained constant. Additional cuts (1302, 1304) are introduced, without introducing corresponding steps.

Abstract: A method including obtaining measurement results of a device manufacturing process or a product thereof, obtaining sets of one or more values of one or more parameters of a distribution by fitting the distribution against the measurement results, respectively, and obtaining, using a computer, a set of one or more values of one or more hyperparameters of a hyperdistribution by fitting the hyperdistribution against the sets of values of the parameters.

Abstract: Methods for calibrating metrology apparatuses and determining a parameter of interest are disclosed. In one arrangement, training data is provided that comprises detected representations of scattered radiation detected by each of plural metrology apparatuses. An encoder encodes each detected representation to provide an encoded representation, and a decoder generates a synthetic detected representation from the respective encoded representation. A classifier estimates from which metrology apparatus originates each encoded representation or each synthetic detected representation. The training data is used to simultaneously perform, in an adversarial relationship relative to each other, a first machine learning process involving the encoder or decoder and a second machine learning process involving the classifier.

Abstract: Parameters of a structure (900) are measured by reconstruction from observed diffracted radiation. The method includes the steps: (a) defining a structure model to represent the structure in a two- or three-dimensional model space; (b) using the structure model to simulate interaction of radiation with the structure; and (c) repeating step (b) while varying parameters of the structure model. The structure model is divided into a series of slices (a-f) along at least a first dimension (Z) of the model space. By the division into slices, a sloping face (904, 906) of at least one sub-structure is approximated by a series of steps (904?, 906?) along at least a second dimension of the model space (X). The number of slices may vary dynamically as the parameters vary. The number of steps approximating said sloping face is maintained constant. Additional cuts (1302, 1304) are introduced, without introducing corresponding steps.

Abstract: A method of determining an estimated intensity of radiation scattered by a target illuminated by a radiation source, has the following steps: obtaining and training (402) a library REFLIB of wavelength-dependent reflectivity as a function of the wavelength, target structural parameters and angle of incidence R(?,?,x,y); determining (408) a wide-band library (W-BLIB) of integrals of wavelength-dependent reflectivity R of the target in a Jones framework over a range of radiation source wavelengths ?; training (TRN) (410) the wide-band library; and determining (412), using the trained wide-band library, an estimated intensity (INT) of radiation scattered by the target illuminated by the radiation source.

Abstract: A method including determining one or more statistical features from data obtained from a lithography process, a lithography apparatus, a substrate processed by the lithography process or the lithography apparatus, wherein determining the one or more statistical features does not include reconstructing a characteristic of the lithography process, of the lithography apparatus, or of the substrate.

Abstract: A method including determining one or more statistical features from data obtained from a lithography process, a lithography apparatus, a substrate processed by the lithography process or the lithography apparatus, wherein determining the one or more statistical features does not include reconstructing a characteristic of the lithography process, of the lithography apparatus, or of the substrate.

Abstract: A method including obtaining measurement results of a device manufacturing process or a product thereof, obtaining sets of one or more values of one or more parameters of a distribution by fitting the distribution against the measurement results, respectively, and obtaining, using a computer, a set of one or more values of one or more hyperparameters of a hyperdistribution by fitting the hyperdistribution against the sets of values of the parameters.

Abstract: Parameters of a structure (900) are measured by reconstruction from observed diffracted radiation. The method includes the steps: (a) defining a structure model to represent the structure in a two- or three-dimensional model space; (b) using the structure model to simulate interaction of radiation with the structure; and (c) repeating step (b) while varying parameters of the structure model. The structure model is divided into a series of slices (a-f) along at least a first dimension (Z) of the model space. By the division into slices, a sloping face (904, 906) of at least one sub-structure is approximated by a series of steps (904?, 906?) along at least a second dimension of the model space (X). The number of slices may vary dynamically as the parameters vary. The number of steps approximating said sloping face is maintained constant. Additional cuts (1302, 1304) are introduced, without introducing corresponding steps.

Abstract: A projection operator framework is described to analyze the concept of localized normal-vector fields within field-material interactions in a spectral basis, in isotropic and anisotropic media. Generate a localized normal-vector field n in a region of the structure defined by the material boundary, decomposed into sub-regions with a predefined normal-vector field and possibly corresponding closed-form integrals. Construct a continuous vector field F using the normal-vector field to select continuous components ET and Dn. Localized integration of normal-vector field n over the sub-regions to determine coefficients of, C. Determine components Ex, Ey, Ez of the electromagnetic field by using field-material interaction operator C to operate on vector field F. Calculate electromagnetic scattering properties of the structure using the determined components of the electromagnetic field.

Abstract: Numerical calculation of electromagnetic scattering properties and structural parameters of periodic structures is disclosed. A reflection coefficient has a representation as a bilinear or sesquilinear form. Computations of reflection coefficients and their derivatives for a single outgoing direction can benefit from an adjoint-state variable. Because the linear operator is identical for all angles of incidence that contribute to the same outgoing wave direction, there exists a single adjoint-state variable that generates all reflection coefficients from all incident waves that contribute to the outgoing wave. This adjoint-state variable can be obtained by numerically solving a single linear system, whereas one otherwise would need to solve a number of linear systems equal to the number of angles of incidence.

Abstract: Numerical calculation of electromagnetic scattering properties and structural parameters of periodic structures is disclosed. A reflection coefficient has a representation as a bilinear or sesquilinear form. Computations of reflection coefficients and their derivatives for a single outgoing direction can benefit from an adjoint-state variable. Because the linear operator is identical for all angles of incidence that contribute to the same outgoing wave direction, there exists a single adjoint-state variable that generates all reflection coefficients from all incident waves that contribute to the outgoing wave. This adjoint-state variable can be obtained by numerically solving a single linear system, whereas one otherwise would need to solve a number of linear systems equal to the number of angles of incidence.

Abstract: A projection operator framework is described to analyze the concept of localized normal-vector fields within field-material interactions in a spectral basis, in isotropic and anisotropic media. Generate a localized normal-vector field n in a region of the structure defined by the material boundary, decomposed into sub-regions with a predefined normal-vector field and possibly corresponding closed-form integrals. Construct a continuous vector field F using the normal-vector field to select continuous components ET and Dn. Localized integration of normal-vector field n over the sub-regions to determine coefficients of, C. Determine components Ex, Ey, Ez of the electromagnetic field by using field-material interaction operator C to operate on vector field F. Calculate electromagnetic scattering properties of the structure using the determined components of the electromagnetic field.