Abstract: A secure smart card is described. The smart card can include a processor, a memory and a transceiver. The smart card can communicate with various terminals and store a digital signature and other information on the card. Another terminal can validate the information stored on the smart card using the digital signature. In certain embodiments, the terminal can also validate the information by using a blockchain. The advanced design of the smart card obviates the need for a network connection.

Abstract: A contactless communication medium including an IC module and a plurality of metal plates. Each of the plurality of metal plates has a through hole and a slit. At least two or more metal plates are laminated via an insulating layer. The metal plates include a first metal plate and a second metal plate. The second metal plate extends across the slit of the first metal plate in plan view of the first metal plate and adds electrostatic capacitance between portions of the slit of the first metal plate.

Abstract: A chemically treated, RFID equipped mesh tire label configured to be integrally incorporated within a vulcanized tire and to provide unique identifier(s) and/or other information about the vulcanized tire during and post tire vulcanization, the label comprising: a mesh face layer configured to be adhered to an outer surface of an unvulcanized tire; a mesh backing layer attached to the mesh face layer and adapted to be integrally incorporated in a vulcanized tire after subjecting a green tire to a vulcanization process; and an RFID device affixed between the mesh face and mesh backing layers, the RFID device that is configured to provide unique identifier(s) and/or other information upon being read with an RFID reader during and post tire vulcanization.

Abstract: A Radio Frequency Identification Device (RFID) tag capable of withstanding electrical stress from induction sealing arrangement and at the same provide shielding against the metallic seal required for induction sealing, comprising at least an antenna coil partitioned into at least a main coil and an auxiliary coil. Both the coils are collocated on two separate overlapping or non-overlapping regions of a dielectric substrate without any galvanic connection there between. The main coil is operatively connected with semiconductor chip disposed on said dielectric substrate for necessary RFID activities, while the auxiliary coil operates as a resonator tuned to a certain frequency to provide shielding against the deleterious effect of the metal seal required for induction sealing.

Abstract: Systems and methods can utilize a conformer model to process a data set for various data processing tasks, including, but not limited to, speech recognition, sound separation, protein synthesis determination, video or other image set analysis, and natural language processing. The conformer model can use feed-forward blocks, a self-attention block, and a convolution block to process data to learn global interactions and relative-offset-based local correlations of the input data.

Abstract: Systems and methods are provided to implement classification of objects, based on sensor data regarding the objects, in a manner that addresses variations in the sensor data, including measurement variables among the objects. A system can include one or more processors to retrieve sensor data regarding an object that is at least one of cellular material from one or more cells, nucleic acid material, biological material, or chemical material. The one or more processors can apply the sensor data as input to a classifier to cause the classifier to determine a classification of the object, the classifier configured based on feature data from a first example of object data and a second example of object data associated with at least one of a different time of detection or a different subject than the first example.

Abstract: A feature management system (the “system”) receives information about a new machine learning (ML) model to be trained. The information includes metadata about the new model. The system applies a trained feature prediction model to the information about the new model and metadata about a plurality of features. The feature prediction model is trained to predict a probability that each of the plurality of features should be selected as an input feature for the new model. The feature management system identifies one or more candidate features based on an output probability score of the feature prediction model. The system presents in a user interface a suggestion to use the one or more candidate features with the new model. The system selects at least one candidate feature and causes the new model to be trained using a set of input features, including the selected candidate feature.

Abstract: A system and method for enhancing C*RAM, improving its performance for known applications such as video processing but also making it well suited to low-power implementation of neural nets. The required computing engine is decomposed into banks of enhanced C*RAM each having a SIMD controller, thus allowing operations at several scales simultaneously. Several configurations of suitable controllers are discussed, along with communication structures and enhanced processing elements.

Abstract: A system for sensing a state of a device is provided. The system includes an autoencoder comprising an encoder, a latent subnetwork, and an extended decoder. The encoder encodes each input data point of input data from an input state space into a latent space to produce latent data points and propagates the latent data points with a neural Ordinary Differential Equation (ODE) to estimate an initial point of latent dynamics of the device in the latent space. The latent subnetwork propagates the initial point till a time index of interest using the neural ODE to produce a state of latent dynamics of the device at the time index of interest. The extended decoder decodes the state of latent dynamics of the device into an output state space different from the input state space to produce output data including the state of the device at the time index of interest.

Abstract: A method, system, and computer program product that is configured to: receive an input time series from an external device in a first system, divide the input time series to a set of univariate time subseries, transform the set of univariate time subseries into a univariate prediction result series using a transformer model, concatenate the univariate prediction result series to a multivariate predictive result, and output the multivariate predictive result for providing time series forecasting to a second system.

Abstract: Embodiments of the present disclosure provide enhanced systems and methods for predicting optimal design flow parameters for optimized output targets for physical design synthesis of a given IC design. A Variational Autoencoder (VAE) along with a regression network are trained using a dataset comprising synthesis design construction flows from historical IC designs to provide a training data representation of the dataset constrained to a latent space of the VAE. The system generates feature vectors based on the training data representation of the dataset and updates the feature vectors with initial design characteristics of the given IC design. The system iteratively performs an input gradient search of the updated feature vectors to optimize an objective function of the design targets to identify locally optimal design parameters. The system identifies globally optimal design flow parameters for optimized design targets based on locally optimal design parameters.

Abstract: The technology relates to providing personalized neural network-based models according to user input, which can be generated upon request or otherwise as needed. This may include receiving, by one or more processors of a computing device, input corresponding to a task description. Then the input corresponding to the task description is encoded into a set of text embeddings. Based on this, the system applies mixer prediction to the set of text embeddings to generate a set of mixers and learns a set of basis models according to the set of mixers. The set of basis models are combined to form a single personalized model corresponding to the task description. This personalized model can then be used in video understanding, quality assessment, providing a recommendation, performing a classification, or performing a search.

Abstract: A method of behavior monitoring includes receiving, at a device, sensor data from one or more sensors associated with a monitored asset. The method also includes applying, at the device, a data scaling operation to input data to generate scaled input data for a pre-trained global model. The input data is based on the sensor data. The method further includes providing, at the device, the scaled input data to the pre-trained global model to selectively generate an alert.

Abstract: Systems and methods are provided to implement classification of objects, based on sensor data regarding the objects, without labels assigned to the sensor data. A system can include one or more processors. The one or more processors can retrieve sensor data regarding an object. The one or more processors can apply the sensor data as input to a classification model to cause the classification model to determine a classification of the object. The classification model can be configured based on training data that includes a plurality of clusters generated by dimensionality reduction of example data regarding example objects. At least one cluster of the plurality of clusters can be associated with the classification. The one or more processors can output the classification of the object.

Abstract: The invention relates to a system and method for detecting anomalous system behaviour. The system comprises a plurality of sensors and a trained autoencoder. The method of training comprises: obtaining training data and test data comprising multiple data records for at least one engineering asset which corresponds to the engineering asset whose behaviour is to be classified, wherein the data records comprise a plurality of sensor readings for the engineering asset; fitting the autoencoder to the obtained training data; running the test data through the encoder of the fitted autoencoder to obtain encodings of the test data; generating a plurality of data sets from the obtained encodings, wherein the generated plurality of data sets include under-represented data sets; cloning the fitted autoencoder to create a cloned autoencoder for each of the generated plurality of data sets; and aggregating the cloned autoencoders to form an over-arching autoencoder.

Abstract: A data classification engine includes an interface configured to interface with an input source, where the input source includes sequential data points representative of a time-varying input signal and one or more processors adapted to receive a temporal sequence of data points at a time T, where the one or more processors are further adapted to receive a next sequential data point and facilitate discarding an oldest data point of the temporal sequence at a time T+1. The data classification engine further includes a matrix adapted to align, at time T, to successive temporal portions of the temporal sequence to generate a set of successive outputs from the temporal sequence and one or more memory modules adapted to store the set of successive outputs from the temporal sequence.

Abstract: Artificial intelligence (AI)-based systems and methods for AI application development using codeless creation of AI workflows is disclosed. The system receives request for creating an artificial intelligence (AI)-based workflow from the user device. Further, the system obtains input data from data sources and pre-process the obtained data using AI based pre-processing model. Further, the system identifies plurality of AI and Generative AI service nodes to be executed on the pre-processed data. The system further generates an AI-based workflow by connecting AI and Generative AI service nodes. Further, the system generates a metadata for AI and Generative AI service nodes by executing each of the identified plurality of AI and Generative AI service nodes. The system validates the metadata based on AI-based rules. Furthermore, the system determines actions to be performed on the metadata based on results of validation and performs the set of actions on the AI-based workflow.

Abstract: Systems, computer-implemented methods, and computer program products to facilitate reducing error propagation when combining generative aligners and transition-based aligners are provided. According to an embodiment, a system can comprise a processor that executed components stored in memory. The computer executable components comprise a generative alignment component, an error propagation component, a discriminative parser, and a stochastic oracle policy component. The error propagation component can compute a posterior distribution over one or more hard alignments of parts given a pair of the generative alignment component. The discriminative parser can be trained via the stochastic oracle policy component to reduce error propagation when combining the generative alignment component with the discriminative parser.

Abstract: A method for processing content management system workflows. Systems and subsystems are established for configuring a content management system to implement workflow processes wherein the content management system (CMS) exposes instances of stored content objects to a plurality of user devices through an electronic interface. Further systems and subsystem are established for identifying metadata maintained by the CMS for the stored content objects, and for identifying a generative AI entity (GAIE) to interact with the CMS. On an ongoing basis, the foregoing systems and subsystems carry out steps for (1) forming a GAIE prompt, wherein the GAIE prompt comprises at least a portion of the metadata identified from the CMS for the stored content objects, (2) receiving a response from the GAIE, wherein the response corresponds to the GAIE prompt; and (3) using, by the CMS, the response from the GAIE to implement processing of a content management system workflow.

Abstract: Certain aspects of the present disclosure provide techniques and apparatus for generating a response to an input query using a generative artificial intelligence model. An example method generally includes receiving an input for processing. A prompt representing the received input is generated based on the received input, contextual information associated with the received prompt, and a prompt-generating artificial intelligence model. The generated prompt is output to a generative artificial intelligence model for processing. A response to the generated prompt is received from the generative artificial intelligence model and output as a response to the received input.

Abstract: Identifying shared events across spiking-neural-network data streams with significant stochastic content. The data streams are first subject to cross correlation. If two data streams are completely uncorrelated, the rate of occurrence, of cross-stream spike pairs, is an approximately uniform “r_ind” across all Time Between Events (TBE's). Any shared events create a gradient, where r_ind increases to a rate “r_shr,” for any TBE's?a Time Of Discernment (TOD). A search for the actual TOD (TOD_a) can be accomplished with a conjectured TOD (TOD_c). TOD_c is tested against an exponential decay with its rate set to a conjectured r_ind (r_ind_c). When r_ind_c=actual r_ind, equal ranges (or regions) of values, of exponential decay, represent equal probabilities. Values of TOD_c and r_ind_c are generated (at respective learning rates), until a combination is found where probabilistically equal regions receive statistically equal numbers of cross-stream events. It is then known TOD_a?TOD_c.

Abstract: The application provides a panoramic perception method, system and a non-transitory computer readable medium. The panoramic perception method comprises: performing a first pretraining on a plurality of weights of a training model using the source database; performing a second pretraining with data augmentation on the plurality of weights of the training model using the source database; performing a combined training on the plurality of weights of the training model using both the source database and the target database; performing a quantization-aware training on the plurality of weights of the training model using the source database and the target database; performing a post training quantization on the plurality of weights of the training model using the target database; and performing panoramic perception by the training model.

Abstract: Provided are a method and apparatus for processing a convolution operation in a neural network, the method includes determining a precision of feature map operands and a precision of weight operands, respectively, on which the convolution operation is to be performed in parallel, decomposing a multiplier included in a convolution operator into sub-multipliers based on the precision of the feature map operands and the precision of the weight operands, performing the convolution operation between the feature map operands and the weight operands by using the decomposed sub-multipliers, each operand being processed in a sub-multiplier corresponding to a precision of the operand, and obtaining output feature maps corresponding to results of the convolution operation.

Abstract: A computer-implemented method for automated handling of data drift in a machine learning (ML) system including a plurality of trained ML models is provided.

Abstract: An embodiment for compressing media utilizing a generative adversarial network (GAN) is provided. The embodiment may include receiving one or more media assets and historical data from a knowledge corpus in accordance with an identified usage context. The embodiment may also include identifying one or more objects in the one or more media assets. The embodiment may further include deriving a relevance score for each identified object. The embodiment may also include creating a training data set. The embodiment may further include applying one or more modifications to each object in a first set. The embodiment may also include in response to determining a GAN discriminator is able to identify each object in the first set modified by the GAN generator as real, generating one or more updated media assets including a second set of one or more objects that are identified by the GAN discriminator as real.

Abstract: A vehicle behavior system comprises a computer system, an observation processor, and neural networks. The observation processor and the neural networks are located in the computer system. The observation processor is configured to receive observations for a vehicle system. The observations are for a current time. The observation processor is configured to extract features from the observations. The neural networks are configured to receive the features extracted from the observations and estimate a behavior for the vehicle system for time steps in response to receiving features extracted from the observations processed by the observation processor. Each of the neural networks is trained to estimate the behavior for the vehicle system for a different time step in the time steps.

Abstract: A method to perform a maintenance operation of an equipment structure is discloses. The method includes collecting monitoring data of the equipment structure disposed in a region of interest, processing the monitoring data to generate normalized monitoring data of the equipment structure, training, based on a first historical portion of the normalized monitoring data as training data and using a machine learning algorithm, an artificial neural network (ANN) model, validating, based on a second historical portion of the normalized monitoring data as validation data, the ANN model to generate a validated ANN model, detecting, using a real time portion of the normalized monitoring data as input to the validated ANN model, an anomaly of the equipment structure, and performing, in response to detecting the anomaly, the maintenance operation of the equipment structure.

Abstract: Methods, systems, and computer program products for managing interactions between a content management system (CMS) and a large language model (LLM) system. The semantics of user questions can be considered before prompting an LLM, or alternatively, before querying datasets that are local to the CMS. Given a user question to be answered, the embedding of the user question can be matched against preconfigured sample question embeddings to determine a best match. A prompt corresponding to the determined best match is then configured based on identification of the class or classes that correspond to the matched question. Prompts for provision to LLMs can be synthesized based on a particular user's identity and/or based on the particular user's historical collaboration activities over objects of the CMS. The LLM can be hosted by a third-party provider. Alternatively all or portions of a large language model system can be hosted within the CMS.

Abstract: According to one embodiment, an information processing device includes one or more processors. The one or more processors are configured to generate a generation function for generating a local waveform used for a continuous wavelet transform, at least part of the generation function being expressed by a machine learning model; and learn the generation function by using learning data including a first input signal.

Abstract: The present application discloses a certifiable out-of-distribution generalization method, a medium, and an electronic device. The method comprises: approximating a deep neural network model using kernelized linear regression; subjecting the deep neural network model to stochastic perturbation learning to derive a classifier for sample separation; and determining a generalization set and certified precision of the deep neural network model, wherein the deep neural network model can output accurate predictions when the perturbation range of semantic information lies within the generalization set, the semantic information being defined as the representation of cascaded intermediate layers of the deep neural network model.

Abstract: A method for training at least one model able to predict a power consumption or production of at least one electric equipment, also called target.

Abstract: A method includes defining a plurality of different windows of time in a recurrent artificial neural network, wherein each of the different windows has different durations, has different start times, or has both different durations and different start times, identifying occurrences of topological patterns of activity in the recurrent artificial neural network in the different windows of time, comparing the occurrences of the topological patterns of activity in the different windows, and classifying, based on a result of the comparison, a first decision that is represented by a first topological pattern of activity that occurs in a first of the windows as less robust than a second decision that is represented by a second topological pattern of activity that occurs in a second of the windows.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for asynchronous deep reinforcement learning. One of the systems includes a plurality of workers, wherein each worker is configured to operate independently of each other worker, and wherein each worker is associated with a respective actor that interacts with a respective replica of the environment during the training of the deep neural network.

Abstract: The present invention relates to an auditing system for a built environment of an age-friendly street based on multisource big data. The auditing system includes: a data acquisition module is configured to acquire urban streetscape image data, urban road network data and urban point-of-interest data; a data classification auditing module is configured to acquire the data of the data acquisition module, classify the image data, and process the image data by using a data processing method to acquire evaluated numerical values of different types of indexes; a data summary analysis module is configured to acquire the evaluated numerical values of the data classification auditing module, calculate sub-item index numerical values of each output unit and calculate result data according to the sub-item index numerical values; a audit result output module is configured to acquire the result data of the data summary analysis module and visualize and output the result data.

Abstract: An information processing device includes: an input unit that inputs a plurality of pieces of time-series data measured by a respective plurality of sensors at different positions; and a processing unit that divides the plurality of pieces of the time-series data into a plurality of pieces of partial time-series data at predetermined time intervals, generates a plurality of primary capsules each including a feature vector of each of the plurality of pieces of the partial time-series data regarding the plurality of pieces of the time-series data, performs graph modeling to generate a weighting matrix in which a connection relationship between the plurality of primary capsules is indicated by a weight corresponding to a distance between the plurality of sensors and the predetermined time interval, and performs graph Fourier transform on the feature vector of each of the plurality of primary capsules based on the weighting matrix.

Abstract: A network learning machine and methods including worker computers that receive instruction communications from assignment computers and an analysis computer that produces training data and creates a network machine learning model that includes at least one parameter and a criterion for optimality, and adjusts the at least one parameter of the machine learning model toward the criterion to optimality based on the training data.

Abstract: Systems and methods for detecting cracks in a surface by analyzing a video, including a full-HD video, of the surface. The video contains successive frames, wherein individual frames of overlapping consecutive pairs of the successive frames have overlapping areas and a crack that appears in a first individual frame of a consecutive pair of the successive frames also appears in at least a second individual frame of the consecutive pair. A fully convolutional network (FCN) architecture implemented on a processing device is then used to analyze at least some of the individual frames of the video to generate crack score maps for the individual frames, and a parametric data fusion scheme implemented on a processing device is used to fuse crack scores of the crack score maps of the individual frames to identify cracks in the individual frames.

Abstract: A model training method includes: acquiring a sample data set corresponding to a target task, a teacher model and an ith initial student model; performing an ith time of channel pruning on the ith initial student model, to acquire a student model subjected to the ith time of channel pruning; performing knowledge distillation according to the sample data set, the teacher model and the student model subjected to the ith time of channel pruning, to acquire an (i+1)th initial student model, wherein a compression ratio of the (i+1)th initial student model to the ith initial student model is equal to a preset ith compression ratio; and updating i to be i+1, and returning to the step of performing the ith time of channel pruning on the ith initial student model, until the updated i is greater than a threshold value N, to acquire a target student model.

Abstract: A system and a method are disclosed for tuning parameters of a large language model. The method comprises identifying first weights of a machine learning (ML) model. Second weights are received from a client device. The second weights may be based on updating, by the client device, the first weights. An update matrix may be generated based on the second weights. The update matrix may be decomposed into first decomposition matrices. Singular values that satisfy a criterion may be identified based on the first decomposition matrices. Singular vectors may be identified based on the singular values. Second decomposition matrices may be identified based on the singular vectors. Updates may be received from the client device of third weights associated with the second decomposition matrices. An updated ML model may be generated based on the updates of the third weights. An inference may be generated based on the updated ML model.

Abstract: The invention discloses a big data analysis system for engine quality detection and prediction, comprising an oil acquisition module for collecting oil in an engine; an oil analysis module for obtaining spectral data, ferrographic data, and physicochemical data of the oil; a data fusion module for fusing the spectral data, ferrographic data and physicochemical data based on a fuzzy logic and a D-S evidence theory to obtain oil fusion data; an oil prediction module for constructing an oil prediction model, training the oil prediction model based on the oil fusion data, and predicting the oil in the engine based on a trained oil prediction model to obtain oil prediction data; a quality detection module connected with the oil prediction module for obtaining a wear degree of the engine and completing a quality prediction of the engine based on the oil prediction data.

Abstract: A counting apparatus includes: a storage unit storing a learning model, the learning model being trained using multiple pairs of training input images each obtained by capturing an image of multiple count target objects with the same shape, and training output images each containing teaching figures that are arranged at respective positions of the multiple count target objects. A captured image acquiring unit acquiring a captured image of multiple count target objects; an output image acquiring unit acquiring an output image in which the count target objects contained in the captured image are converted into count target figures, by applying the captured image to the learning model. A counting unit counting the number of count target objects, using the multiple count target figures contained in the output image; and an output unit outputting the number of count target objects counted by the counting unit.

Abstract: Provided is a computer-implemented method for generating a machine learning model to classify an account based on merchant activation, including providing an input to a generator network of a generative adversarial network (GAN) to generate an output; providing the output as input to a discriminator network; providing a training dataset as input to the discriminator network; and updating the generator network based on a first output of the discriminator network having a label that indicates whether a set of values of each of the plurality of features is a real set of values or a fake set of values. The method may include updating the discriminator network based on a second output of the discriminator network having a label that indicates whether a selected account of the plurality of accounts is going to conduct a first payment transaction. A system and computer program product are also provided.

Abstract: Disclosed in the present disclosure are a transfer reinforcement learning method and apparatus, multi-task reinforcement learning method and apparatus, relating to the field of intelligent control technology. The transfer reinforcement learning method includes determining operational instructions for instructing an agent to perform a first task; determining an inclusion relation between multiple second tasks and the first tasks based on the operational instructions; determining a shared parameter set corresponding to the multiple second tasks based on the inclusion relation between the multiple second tasks and the first task, wherein the shared parameter set includes a plurality of parameters shared by the multiple second tasks; and performing transfer reinforcement learning based on the shared parameter set and the first task to obtain model parameters of a target policy model corresponding to the first task.

Abstract: Exemplary systems, methods, and computer-accessible medium are provided that can train a language model for a medical use or performing a medically-related procedure. Thus, exemplary systems, methods, and computer-accessible medium can be provided that can model a reward neural network on one or more physician preferences and train the language model by applying the reward neural network modeled on the physician preference(s) as feedback to guide the language model to learn the physician preference(s). The reward neural network can rely on an artificial intelligent (AI) model as a surrogate reward function for a physician feedback, and can obtain the physician preference(s) implicitly from electronic health records and/or other sources of medical data.

Abstract: A method is provided that includes: obtaining a first Text-to-Image model and a pre-trained reward model, wherein the first Text-to-Image model is used to generate a corresponding image based on input text, and the pre-trained reward model is used to score a data pair composed of the input text and the corresponding generated image; and adjusting the parameters of the first Text-to-Image model based on the pre-trained reward model and a reinforcement learning policy to obtain a second Text-to-Image model.

Abstract: A system and a method are disclosed, the method including receiving, by a first local controller of a first edge device, an input associated with an environment in which the first edge device operates, using a first machine-learning algorithm, determining, by the first local controller, a parameter for a pre-trained modem algorithm of the first edge device based on the input, executing a task on the first edge device based on executing the pre-trained modem algorithm with the parameter, determining a result of executing the task, training the first machine-learning algorithm, generating a first update to the first machine-learning algorithm based on the training, sending the first update to a server, receiving, from the server, a server update to the first machine-learning algorithm, and based on the server update, updating the first machine-learning algorithm.

Abstract: A computer implemented method is disclosed for facilitating execution of a Machine Learning (ML) model by a system of resource nodes, the ML model comprising a plurality of functional model parts. The method, performed by a resource node of the system, comprises generating a placement map for the ML model, wherein the placement map specifies, for each of the functional model parts, a mapping between the functional model part and at least one resource node of the system that is to execute the functional model part. The method further comprises identifying, from the placement map, a functional model part that is to be executed by the resource node, and executing the identified functional model part.

Abstract: A computer-implemented method for operating a technical device using a model based on artificial intelligence by a client of a client-server system, wherein each client stores a client model for operating a technical device, where a global model is provided to a client based on federated learning, and where the technical device is operated using the client model and sensitive parameter.

Abstract: Methods, systems, and computer program products for managing interactions between a content management system (CMS) and a large language model (LLM) system. The semantics of user questions can be considered before prompting an LLM, or alternatively, before querying datasets that are local to the CMS. Given a user question to be answered, the embedding of the user question can be matched against preconfigured sample question embeddings to determine a best match. A prompt corresponding to the determined best match is then configured based on identification of the class or classes that correspond to the matched question. Prompts for provision to LLMs can be synthesized based on a particular user's identity and/or based on the particular user's historical collaboration activities over objects of the CMS. The LLM can be hosted by a third-party provider. Alternatively all or portions of a large language model system can be hosted within the CMS.

Abstract: A system includes an agent engine, an encoder, a general-purpose solver engine, and an orchestrator. The orchestrator is configured to receive a first problem instance corresponding to a learned policy that is based on auto reinforcement learning, and provide the first problem instance to the general-purpose solver engine, which is configured to execute based on the first problem instance to determine a solver state. The orchestrator is configured to extract, from the general-purpose solver engine, the solver state, and to provide the solver state to the encoder. The encoder is configured to query the agent engine for a best action according to the learned policy and an encoded solver state. The agent engine is configured to determine the best action according to the learned policy and the encoded solver state. The orchestrator is configured to receive the best action, and direct the general-purpose solver to implement the best action.