Patents.

Apparatus, method and storage medium for database query

An apparatus, method and storage medium for database query: determine that a syntax tree corresponding to an SQL query statement includes a preset subtree indicating a query mode for querying vector data; determine a query mode according to the preset subtree and the SQL statement; query data in a database based on the query mode. Enables efficient approximate vector query from SQL instead of an API.

Generative Model Based Decomposition of Input Query into Sub-Queries and Generation of Response

Multi-dimensional periodicity detection of IoT device behavior

Skip-list checkpoint creation

Indexing structured data with security information

A system for indexing and searching includes an input interface and a processor. The interface is to receive a request to search for a term. The processor is to determine a search response based at least in part on a security associated with an index field and the term.

Regulatory sequences

The present disclosure provides modular regulatory elements useful for transcription and gene expression.

Managed Metastorage

The present application discloses a method, system, and computer system for providing access to information stored on system for data storage. The method includes receiving a data request from a user, determining data corresponding to the data request, determining whether the user has requisite permissions to access the data, and in response to determining that the user has requisite permissions to access the data: determining a manner by which to provide access to the data, wherein the data comprises a filtered subset of stored data, and generating a token based at least in part on the user and the manner by which access to the data is to be provided.

Autonomous vehicle platform and safety architecture

In embodiments of an autonomous vehicle platform and safety architecture, safety managers of a safety-critical system monitor outputs of linked components of the safety-critical system. The linked components comprise at least three components, each of which is configured to produce output indicative of a same event independent from the other linked components by using different input information than the other linked components. The safety managers also compare the outputs of the linked components to determine whether each output indicates the occurrence of a same event. When the output of one linked component does not indicate the occurrence of an event that is indicated by the outputs of the other linked components, the safety managers identify the one linked component as having failed. Based on this, the outputs of the other linked components are used to carry out operations of the safety-critical system without using the output of the failed component.

Methods and Systems for Inferring Gene Expression Using Cell-Free DNA Fragments

Defending against adversarial attacks in federated learning

A computer-implemented method, a computer program product, and a computer system for defending against adversarial attacks in federated learning. In the federated learning comprising an aggregator and parties, the aggregator receives weights sent from the respective parties. The aggregator computes values of a performance metric for weight arrays obtained by the respective parties, using a validation dataset. The aggregator ranks the values of the performance metric in a list. The aggregator recursively splits the list in half until one or more adversary updates of the weights are isolated. The aggregator excludes one or more parties that send the one or more adversary updates from participating in a current round of training in the federated learning.

Search using an overlay graph mapping to source knowledge graphs

Provided are techniques for a search using an overlay graph mapping to source knowledge graphs. A plurality of overlay graphs are generated, where each overlay graph comprises entities represented by nodes and relations represented by edges, and where the entities and the relations map to a subset of entities and relations in a plurality of source knowledge graphs. A search request comprising an entity and a relation is received. An overlay graph is selected from the plurality of overlay graphs based on the entity and the relation. The search request is issued against the overlay graph, where the search request is translated to knowledge graph specific queries, and where the knowledge graph specific queries are issued against the plurality of source knowledge graphs. Search results are received from the plurality of source knowledge graphs. The search results are used to respond to the search request.

Autonomous data sharding and topology alterations for disaster recovery preparation

A computer-implemented method, a computer program product, and a computer system for data sharding and topology alterations for disaster recovery preparation. A computer system estimates time needed to copy a data store of a data center that is assumed offline. A computer system determines whether ...

Parallel join application for machine learning features

A technique for generating a value for a machine learning feature: receiving a machine learning feature and a context, generating a plurality of distinct sub-contexts based on the context, and assigning each of the plurality of sub-contexts to a parallel process.

Systems and methods for online adaptation for cross-domain streaming data

Embodiments described herein provide an online domain adaptation framework based on cross-domain bootstrapping for online domain adaptation, in which the target domain streaming data is deleted immediately after adapted. At each online query, the data diversity is increased across domains by bootstrapping the source domain to form diverse combinations with the current target query. To fully take advantage of the valuable discrepancies among the diverse combinations, a set of independent learners are trained to preserve the differences. The knowledge of the learners is then integrated by exchanging their predicted pseudo-labels on the current target query to co-supervise the learning on the target domain, but without sharing the weights to maintain the learners' divergence.

Systems and methods for generating improved decision trees

A system and method for generating a decision tree having a plurality of nodes, arranged hierarchically as parent nodes and child nodes. Generating a node includes receiving training data with data instances having attributes and labels, instance weightings, a valid domain for each attribute, and an accuracy criterion.

System and method of providing a platform for managing data content campaign on social networks

A computer-implemented system including a data store: a content items database, a user account database, and one or more servers configured to execute a computer program to perform one or more of the following: receiving content items from content database that are associated with a topic selected by a user for posting on a social network, wherein at least one content item is associated with an URL; estimating a post to a reaction filter for a time interval for the social network for the user, calculating a reaction profile associated with reactions to posts on the social network by aggregating reaction time of a plurality of users on the social network for one or more content items posted on the social network; determining a schedule for posting the content items on the social network as a function of the post to reaction filter and reaction profile.

Display Screen or Portion Thereof with Transitional Graphical User Interface

Generating dual sequence inferences using a neural network model

Hardware secure enclave and blockchain based system and method for securing and monetising access to data

Systems and methods for video-language neural networks

Embodiments described herein provide a vision-language neural network framework that outputs a text response to a user text query relating to the media content of the video input. Specifically, the vision-language neural network may comprise (1) a vision encoder (ViT) transforming each frame input from the video input into a set of tokens, (2) a frame-level tokenizer to reduce the number of tokens, (3) a temporal encoder to build video-level token representations, and (4) an autoregressive LLM generating a text output based on such video tokens and text prompt tokens.

Systems and methods for training and inference of large multimodal models

Embodiments described herein provide a method of performing a vision-language task by a neural network multimodal model in response to multiple input images, the method comprising: receiving, via a data interface, a text input and an image input; generating text tokens based on the text input; generating a plurality of image patches, wherein each image patch of the plurality of image patches includes a portion of the image input at substantially a same resolution as the image input; generating a downsized image based on a downsizing of the image input; generating vision tokens based on the plurality of image patches and the downsized image; generating, via a neural network based language model, an output based on the text tokens and the vision tokens; and updating parameters of the neural network based language model based on a loss objective based on the output.

Optimizing operator configuration in containerized environments

Computer technology for optimizing operator configuration for deployment of an application in a container orchestration cluster. The computer technology includes machine logic for collecting historical data for deployment data including operator configurations, cluster configurations, and behavior data.

Differentially private query budget refunding

A differentially private security system communicatively coupled to a database storing restricted data receives a database query from a client. The database query includes a relation specifying a set of data in the database upon which to perform the query and privacy parameters associated with the query.

Quasi-recurrent Neural Network Based Encoder-decoder Model

The technology disclosed provides a quasi-recurrent neural network (QRNN) encoder-decoder model that alternates convolutional layers, which apply in parallel across timesteps, and minimalist recurrent pooling layers that apply in parallel across feature dimensions.

Secured search for ready-made search software

Search management systems and methods that cryptographically-secure search indices, search queries, and associated document records while in-use by cloud-based search software without requiring modification to the cloud-based search core software. The search proxy resides between a cloud-based search service and its clients. (abstract excerpt from Google Patents)

Private vertical federated learning

A second set of data identifiers, comprising identifiers of data usable in federated model training by a second data owner, is received at a first data owner from the second data owner. An intersection set of data identifiers is determined at the first data owner. At the first data owner according to the intersection set of data identifiers, the data usable in federated model training is rearranged by the first data owner to result in a first training dataset. At the first data owner using the intersection set of data identifiers, the first training dataset, and a previous iteration of an aggregated set of model weights, a first partial set of model weights is computed. An updated aggregated set of model weights, comprising the first partial set of model weights and a second partial set of model weights from the second data owner, is received from an aggregator.

Systems and methods for a text-to-video generation framework

Embodiments described herein provide a generation model comprising a video-specific variational auto-encoder (VAE) for effective compression of video pixel information with reduced spatial and temporal dimensions and a video diffusion transformer (vDiT) to generate latent representations of frames. Specifically, the VAE may, instead of encoding each frame independently, incorporate both temporal and spatial compression. This significantly decreases the token length, improves the computational cost of training and inference, and facilitates the generation of long videos. The encoded training video, in the form of latent representations from a VAE encoder may then be passed to the vDiT to reconstruct the latent representations during training. The trained vDiT may then generate latent representations of a video in response to a text input, and the latent representations may be converted to a video output by a VAE decoder.

Content based image retrieval for lesion analysis

Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are commonly used to assess patients with known or suspected pathologies of the lungs and liver. In particular, identification and quantification of possibly malignant regions identified in these high-resolution images is essential for accurate and timely diagnosis. However, careful quantitative assessment of lung and liver lesions is tedious and time consuming. This disclosure describes an automated end-to-end pipeline for accurate lesion detection and segmentation.

Training, testing, and verifying autonomous machines using simulated environments

In various examples, physical sensor data may be generated by a vehicle in a real-world environment. The physical sensor data may be used to train deep neural networks (DNNs). The DNNs may then be tested in a simulated environment - in some examples using hardware configured for installation in a vehicle - to control a virtual vehicle in the simulated environment.

Data sharing for network connected systems

The present application discloses a method, system, and computer system for providing access to data. The method includes receiving, by a data manager service from a data requesting service, a request using an identifier for a high-level data object to access a set of data associated with the high-level data object, determining, by the data manager service, low-level data object(s) corresponding to the set of data based on the identifier for the high-level data object, determining whether a user associated with the request has permission to access at least a subset of the low-level data object(s), and in response to determining that the user associated has permission to access the at least the subset of the low-level data object(s), generating, by the data manager service, a uniform resource locator (URL) via which the at least the subset of the one or more low-level data objects is accessible by the user.

Byzantine-robust federated learning

A federated learning method comprises creating a log of previously provided gradients from a plurality of workers, receiving updated gradients from the plurality of workers, calculating a vulnerability weight for each layer of a global machine learning model using the updated gradients, calculating an aggregated gradient using the vulnerability weight and the updated gradients, and updating the global machine learning model using the aggregated gradient. Some embodiments may also determine whether a Byzantine attack is occurring based upon the calculated aggregated gradient. An apparatus and computer program product may be used to implement the method.

Delivering Domain-Expert Agents and Models Using Synthetic Knowledge

Scaling machine learning using dynamic sharding

The system includes one or more processors configured to determine that a model is to be updated; determine a shard on which the model is to be deployed; determine whether to move the model to a different shard; in response to determining that the model is to be moved to the different shard, allocate the model to the different shard; and restart the different shard.

Systems and methods for training and evaluating multimodal neural network based language models

Embodiments described herein provide a method of building an artificial intelligence (AI) agent to respond to a task request from a user. The method includes: receiving a set of single-modal data samples of a plurality of modalities; selecting a first single-modal data sample of a first modality and a second single-modal data sample of a second modality; generating a question associated with the first single-modal data sample and the second single-modal data sample; generating an answer with a reasoning to the question based on a second input prompt; training, a second neural network based language model, using a dataset comprising the question and the answer to generate a candidate answer in response to a training query; building the AI conversation bot through an application programming interface to the trained second neural network language model; and generating, using the AI conversation bot, a response to the task request.

Analytics driven engagement

A computerized method performed by a server computer comprising receiving and storing a set of engagement rules for a website based on previous analytics, wherein each engagement rule defines criteria based on user interaction with the website.

Privacy-preserving federated learning

A plurality of public encryption keys are distributed to a plurality of participants in a federated learning system, and a first plurality of responses is received from the plurality of participants, where each respective response of the first plurality of responses was generated based on training data local to a respective participant of the plurality of participants and is encrypted using a respective public encryption key of the plurality of public encryption keys. A first aggregation vector is generated based on the first plurality of responses, and a first private encryption key is retrieved using the first aggregation vector. An aggregated model is then generated based on the first private encryption key and the first plurality of responses.

System and Method for Identifying Sentiment in Text Strings

Data sharing for network connected systems

The present application discloses a method, system, and computer system for providing access to data. The method includes receiving, by a data manager service from a data requesting service, a request using an identifier for a high-level data object to access a set of data associated with the high-level data object, determining, by the data manager service, low-level data object(s) corresponding to the set of data based on the identifier for the high-level data object, determining whether a user associated with the request has permission to access at least a subset of the low-level data object(s), and in response to determining that the user associated has permission to access the at least the subset of the low-level data object(s), generating, by the data manager service, a uniform resource locator (URL) via which the at least the subset of the one or more low-level data objects is accessible by the user.

Delivering Domain-Expert Agents for Improving Problem-Solving

Merge operations for darts

A one-shot neural architecture search method referred to as MergeNAS by merging different types of convolutions into a single operation. This mergence approach reduces the search cost to roughly half a GPU-day as well as alleviates the over-fitting problem caused by a traditional differentiable architecture search (DARTS) approach by reducing the number of redundant parameters.

Aggregation operations in a distributed database

Detecting handheld device movements utilizing a handheld-movement-detection model

Federated learning technique for applied machine learning

A method, a computer program product, and a system of training a machine learning model using federated learning with extreme gradient boosting. The method includes computing an epsilon hyperparameter using training dataset sizes from a first party and a second party. The method also includes transmitting a machine learning model and the epsilon hyperparameter to the first party and the second party and receiving a first model update and a second model update from the first party and the second party respectively. The method further includes fusing the first model update and the second model update to produce a global histogram and determining at least one split candidate in a decision tree used by the machine learning model using the global histogram. The method also includes rebuilding the machine learning model by adding the split candidate to a decision tree of the machine learning model.

Efficiently serving machine-learned model computations with high throughput and low latency

An example method includes receiving input requests to process a plurality of input sequences using the machine-learned sequence processing model to generate a plurality of output sequences respectively corresponding to the plurality of input sequences; generating a plurality of initial attention tensors respectively for the plurality of input sequences, wherein: one or more respective initial attention tensors are generated for each respective input sequence in parallel over input elements of the respective input sequence; and the one or more respective initial attention tensors are generated in one or more batches having a first batch size using a prefill system that comprises one or more prefill computing devices and executes one or more layers of the machine-learned sequence processing model; and autoregressively generating, using the plurality of initial attention tensors, a plurality of output elements for each of the plurality of output sequences in one or more batches having a second batch size, wherein: the plurality of output elements are generated using a generation system that comprises one or more generation computing devices.

Enterprise retrieval-augmented generation system

An enterprise data source contains documents and identifiers for an enterprise. A RAG data ingestion platform retrieves a document and document identifier from the enterprise data source and divides the document into a first set of chunks. A first LLM query, designed to predict questions associated with the retrieved document based on the first set, is output to a first LLM. The platform executes a first embedding model on a response to the first LLM query and document metadata including the document identifier and stores a result of the first embedding model in the RAG vector database. The retrieved document is also divided into a second set of chunks (with chunks smaller than the first set and including a second chunk identifier). A second embedding model is executed based on the second set of chunks, and a result of the second embedding model is stored in the RAG vector database.

Method and apparatus for lightweight and parallelization of accelerator task scheduling

Disclosed are a method and an electronic apparatus including an accelerator for lightweight and parallel accelerator task scheduling. The method includes pre-running a deep learning model with sample input data having a preset data form and generating a scheduling result through the pre-running.

Automatic verification of decentrailized protocols

Systems and methods for non-invasive blood pressure measurement

Systems and methods for non-invasive blood pressure measurement are disclosed, using a wearable member generating PPG signals and a blood pressure calculation system that computes an arterial blood pressure value from extracted feature vectors.

Delegable Automated System for Review of Network Threats

Event intensity assessment

Event intensity assessment can include detecting an event description within textual input received via a data communication network. An event-correlated data structure based on the event can be generated, the event-correlated data structure including an event descriptor corresponding to the event. An event sentiment can be determined based on the event descriptor and an event impact based on a quantitative temporal-spatial measure corresponding to the event.