Abstract: The technology disclosed provides systems and methods related to preventing exfiltration of training data by feature reconstruction attacks on model instances trained on the training data during a training job. The system comprises a privacy interface that presents a plurality of modulators for a plurality of training parameters. The modulators are configured to respond to selection commands via the privacy interface to trigger procedural calls. The procedural calls modify corresponding training parameters in the plurality of training parameters for respective training cycles in the training job. The system comprises a trainer configured to execute the training cycles in dependence on the modified training parameters. The trainer can determine a performance accuracy of the model instances for each of the executed training cycles.

Abstract: A method for making a component for use in a plasma processing chamber is provided. A non-oxide silicon containing powder composition is placed in a mold, wherein the non-oxide silicon containing powder composition consists essentially of a non-oxide silicon containing powder and at least one of a B or B4C dopant. The non-oxide silicon containing powder composition is subjected to spark plasma sintering (SPS) to form a spark plasma sintered component. The spark plasma sintered component is machined into a plasma processing chamber component.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

Abstract: A component of a plasma processing chamber having a coating on at least one surface that comprises yttrium aluminum. The coating is an aerosol deposited coating from a powder mixture of an yttrium oxide powder and an aluminum-containing powder and having an yttrium to aluminum ratio of 4:1 to 1:4 by molar number. The coating can be annealed to form a porous ternary oxide.

Abstract: A method for making a component for use in a semiconductor processing chamber is provided. A component body is formed from a conductive material having a coefficient of thermal expansion of less than 10.0×10?6/K. A metal oxide layer is then disposed over a surface of the component body.

Abstract: A method for forming a coating on a component of a substrate processing system includes arranging the component in a processing chamber and applying a ceramic material to form the coating on one or more surfaces of the component. The ceramic material is comprised of a mixture including a rare earth oxide and having a grain size of less than 150 nm and is applied while a temperature within the processing chamber is less than 400° C. The coating has a thickness of less than 30 ?m. A heat treatment process is performed on the coated component in a heat treatment chamber. The heat treatment process includes increasing a temperature of the heat treatment chamber from a first temperature to a second temperature that does not exceed a melting temperature of the mixture over a first period and maintaining the second temperature for a second period.

Abstract: A component of a plasma processing chamber having at least one plasma facing surface of the component comprises single crystal metal oxide material. The component can be machined from a single crystal metal oxide ingot. Suitable single crystal metal oxides include spinel, yttrium oxide, and yttrium aluminum garnet (YAG). A single crystal metal oxide can be machined to form a gas injector of a plasma processing chamber.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic biometric identifiers. The method includes generating a scannable code upon receiving a success nonce from a registration server. The registration server can access a user identifier and a hash of at least a signature using the success nonce. The signature can be generated based at least in part upon a biometric identifier of a user. The method includes recreating the hash of the signature stored by the registration server. The method includes generating the scannable code by encrypting the success nonce and the recreated hash. The biometric identifier of the user is generated by feeding a plurality of non-deterministic biometric inputs to a trained machine learning model producing a plurality of feature vectors. The method includes projecting the plurality of feature vectors onto a surface of a unit hyper-sphere and computing a characteristic identity vector representing the user.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

Abstract: The present application provides a radial counter flow jet gas cooling system for a rotor of a dynamoelectric machine. The radial counter flow jet gas cooling system may include a centering pin, a number of axial inlet ducts, a number of radial outlet ducts in communication with the axial inlet ducts, an axial subslot positioned about the axial inlet ducts, and a radial counter flow duct in communication with the axial subslot and extending along the centering pin.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

Abstract: An apparatus adapted for use in a plasma processing chamber is provided. An aluminum body with at least one surface is provided. An aluminum oxide containing aerosol deposition coating is disposed over the at least one surface of the aluminum body. An yttrium containing aerosol deposition coating is disposed over the aluminum oxide containing aerosol deposition coating.

Abstract: Textured silicon components of a semiconductor processing chamber having hillock-shaped or pyramid-shaped structures on its surface, and a method of texturing such silicon components. The silicon component can be selectively textured using chemical means to form the hillock-shaped structures to increase the surface area of the silicon component to improve polymer adhesion.

Abstract: The technology disclosed relates to a system and method for detecting when to diverge to a tangential conversation from a linearly structured steps of a cognitive behavioral therapy (CBT) session in a series of sessions for a mental health disorder. The system includes logic to determine when to merge the tangential conversation back to flow of linearly structured steps in the therapy session. The system can bookmark the last completed step in the flow of linearly structured steps of the CBT therapy session before starting the tangential conversation. The system can calculate a return to CBT session score after each succeeding utterance from the patient in the tangential conversation. The system resumes flow of the linearly structured steps of the CBT session upon detecting that the return to CBT session score has exceeded a threshold.

Abstract: The technology disclosed relates to a system and method for remembering content received from a patient for future use and understood by computerized natural language processing during a chatbot therapy session. The system includes a graph of entities and relationships in an artificial memory graph data structure. The entity nodes in the graph represent entities that are related to a patient root node representing the patient. The entity nodes are connected to patient root node by one or more relationship edges that have named roles. The entity nodes and/or relationship edges can be slotted for at least a distinguishing name, a patient sentiment towards the entity and an entity state.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

Abstract: A component for use as part of a plasma processing chamber for processing a wafer is provided. The component comprises a component body of silicon carbide doped with at least one of tungsten, tantalum, or boron.

Abstract: The technology disclosed provides systems and methods related to preventing exfiltration of training data by feature reconstruction attacks on model instances trained on the training data during a training job. The system comprises a privacy interface that presents a plurality of modulators for a plurality of training parameters. The modulators are configured to respond to selection commands via the privacy interface to trigger procedural calls. The procedural calls modify corresponding training parameters in the plurality of training parameters for respective training cycles in the training job. The system comprises a trainer configured to execute the training cycles in dependence on the modified training parameters. The trainer can determine a performance accuracy of the model instances for each of the executed training cycles.

Abstract: The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.