Abstract: Systems and methods for training a machine learning model are disclosed. A system may be configured to obtain a plurality of training samples. The system includes a machine learning model to generate predictions and generate a confidence score for each generated prediction. In this manner, the system is configured to, for each training sample of the plurality of training samples, generate a prediction by a machine learning model based on the training sample and generating a confidence score associated with the prediction by the machine learning model. The system is also configured to train the machine learning model based on the plurality of predictions and associated confidence scores. For example, one or more training samples may be excluded from use in training the machine learning model based on the associated one or more confidence scores (such as the confidence score being less than a threshold).

Abstract: Certain aspects of the disclosure provide systems and methods for generating meaningful insights from a data frame based on an insight score. An insight score may quantify the significance and confidence of a given insight. Aspects of the disclosure provide for optimizing the most meaningful insight based on a greedy binary search approach. Aspects of the disclosure further provide for obtaining the optimal insight based on a gradient search approach.

Abstract: A first large language model (LLM) instance may be instructed to request data while being prevented from performing calculations using the data. A second LLM instance may be instructed to provide a response to the request for data based on a known complete data set. The response may be translated into a machine-readable response in a format configured for processing by a calculation engine. The calculation engine may process the machine-readable response, thereby generating a calculation engine output. A mismatch between the calculation engine output and a known result obtained using the known complete data set may be identified, and the instruction to the first LLM may be modified in response.

Abstract: Systems and methods are disclosed for detecting hallucinations in large language models (LLMs). An example method includes receiving a first prompt for submission to the first LLM, generating, using the first LLM, a plurality of semantically equivalent prompts to the first prompt, generating, using the first LLM, a first response to the first prompt and a plurality of second responses to the plurality of semantically equivalent prompts, generating, using a second LLM, a plurality of third responses to the semantically equivalent prompts, generating a semantic consistency score for the first response based at least in part on the first prompt, the plurality of semantically equivalent prompts, the plurality of second responses, and the plurality of third responses, and determining whether or not the first response is an accurate response to the first prompt based at least in part on the semantic consistency score.

Abstract: Augmented Denoising Diffusion Implicit Models (“DDIMs”) using a latent trajectory optimization process can be used for image generation and manipulation using text input and one or more source images to create an output image. Noise bias and textual bias inherent in the model representing the image and text input is corrected by correcting trajectories previously determined by the model at each step of a diffusion inversion process by iterating multiple starts the trajectories to find determine augmented trajectories that minimizes loss at each step. The trajectories can be used to determine an augmented noise vector, enabling use of an augmented DDIM and resulting in more accurate, stable, and responsive text-based image manipulation.

Abstract: Certain aspects of the present disclosure provide techniques for executing a function in a software application through a conversational user interface based on a knowledge graph associated with the function. An example method generally includes receiving a request to execute a function in a software application through a conversational user interface. A graph definition of the function is retrieved from a knowledge engine. Input is iteratively requested through the conversational user interface for each parameter of the parameters identified in the graph definition of the function based on a traversal of the graph definition of the function. Based on a completeness graph associated with the function, it is determined that the requested inputs corresponding to the parameters identified in the graph definition of the function have been provided through the conversational user interface. The function is executed using the requested inputs as parameters for executing the function.

Abstract: Aspects of the present disclosure provide techniques for active multifidelity machine learning. Embodiments include selecting, based on one or more criteria, a first subset of unlabeled training data for manual review and a second subset of unlabeled training data for providing to a pre-trained machine learning model for automated labeling. Embodiments include receiving manual label data for the first subset of unlabeled training data. Embodiments include providing inputs to the pre-trained machine learning model based on a subset of the manual label data and the second subset of training data. Embodiments include receiving, as outputs from the pre-trained machine learning model, automated label data for the second subset of unlabeled training data. Embodiments include generating a training data set for a target machine learning model based on the set of unlabeled training data, the manual label data, and the automated label data.

Abstract: A method includes receiving a user input including natural language text. The method also includes generating modified inputs from the user input. The method also includes executing a machine learning model on the modified inputs to generate model outputs. The method also includes sampling the model outputs using a statistical sampling strategy to generate sampled model outputs. The method also includes clustering the sampled model outputs into clusters. The method also includes generating a confidence metric of the clusters. The method also includes routing, automatically in a computing system, the user input based on whether the confidence metric satisfies a threshold value.

Abstract: A method implements a dataset rank metric for measuring dataset relevance. Metadata is identified for a plurality of datasets. A graph structure is generated in storage. The graph structure includes a multitude of nodes connected by a multitude of edges. Each node of the multitude of nodes representing a respective dataset of a multitude of datasets, the multitude of edges connecting the multitude of nodes according to a data lineage determined from metadata of the multitude of datasets. A composite score is generated for each node of the graph. The computer processor iteratively processes the composite scores for the multitude of nodes of the graph to generate a dataset rank for each dataset. The multitude of datasets is presented in an interface, sorted according to the respective dataset rank of each dataset.

Abstract: Augmented Denoising Diffusion Implicit Models (“DDIMs”) using a latent trajectory optimization process can be used for image generation and manipulation using text input and one or more source images to create an output image. Noise bias and textual bias inherent in the model representing the image and text input is corrected by correcting trajectories previously determined by the model at each step of a diffusion inversion process by iterating multiple starts the trajectories to find determine augmented trajectories that minimizes loss at each step. The trajectories can be used to determine an augmented noise vector, enabling use of an augmented DDIM and resulting in more accurate, stable, and responsive text-based image manipulation.

Abstract: Certain aspects of the present disclosure provide techniques for executing a function in a software application through a conversational user interface based on a knowledge graph associated with the function. An example method generally includes receiving a request to execute a function in a software application through a conversational user interface. A graph definition of the function is retrieved from a knowledge engine. Input is iteratively requested through the conversational user interface for each parameter of the parameters identified in the graph definition of the function based on a traversal of the graph definition of the function. Based on a completeness graph associated with the function, it is determined that the requested inputs corresponding to the parameters identified in the graph definition of the function have been provided through the conversational user interface. The function is executed using the requested inputs as parameters for executing the function.

Abstract: Certain aspects of the present disclosure provide techniques for optimizing results generated by functions executed using a rule-based knowledge graph. The method generally includes generating a neural network based on a knowledge graph and inputs for performing a function using the knowledge graph. Inputs for the function are received and used to generate a result of the function. A request to optimize the generated result of the function is received. A loss function is generated for the neural network. Generally, the loss function identifies a desired optimization for the function. Values of parameters in the neural network are adjusted to optimize the generated result based on the generated loss function, and the adjusted values of the parameters in the neural network are output in response to the request to optimize the generated result of the function.

Abstract: Systems and methods for training machine learning models are disclosed. An example method includes receiving historical event timing data including event data for a first portion including events from a first time period, and a second portion comprising events from a second time period not including the first time period, predicting, based on the first portion of the historical event timing data, a first plurality of predicted events, the first plurality of predicted events corresponding to the second time period, determining a first subset of predicted events to be accurate predictions based at least in part on comparing the first plurality of predicted events to the historical events occurring within the second time period, generating training data based at least in part on the first subset of the first plurality of predicted events, and training the machine learning model based at least in part on the training data.

Abstract: Certain aspects of the present disclosure provide techniques for performing finite rank deep kernel learning. In one example, a method for performing finite rank deep kernel learning includes receiving a training dataset; forming a set of embeddings by subjecting the training dataset to a deep neural network; forming, from the set of embeddings, a plurality of dot kernels; linearly combining the plurality of dot kernels to form a composite kernel for a Gaussian process; receiving live data from an application; and predicting a plurality of values and a plurality of uncertainties associated with the plurality of values simultaneously using the composite kernel.

Abstract: Systems and methods are disclosed for model based document image enhancement. Instead of requiring paired dirty and clean images for training a model to clean document images (which may cause privacy concerns), two models are trained on the unpaired images such that only the dirty images are accessed or only the clean images are accessed at one time. One model is a first implicit model to translate the dirty images from a source space to a latent space, and the other model is a second implicit model to translate the images from the latent space to clean images in a target space. The second implicit model is trained based on translating electronic document images in the target space to the latent space. In some implementations, the implicit models are diffusion models, such as denoising diffusion implicit models based on solving ordinary differential equations.

Abstract: A method implements a dataset rank metric for measuring dataset relevance. Metadata is identified for a plurality of datasets. A graph structure is generated in storage. The graph structure includes a multitude of nodes connected by a multitude of edges. Each node of the multitude of nodes representing a respective dataset of a multitude of datasets, the multitude of edges connecting the multitude of nodes according to a data lineage determined from metadata of the multitude of datasets. A composite score is generated for each node of the graph. The computer processor iteratively processes the composite scores for the multitude of nodes of the graph to generate a dataset rank for each dataset. The multitude of datasets is presented in an interface, sorted according to the respective dataset rank of each dataset.

Abstract: Certain aspects of the present disclosure provide techniques for mapping natural language to stored information. The method generally includes receiving a long-tail query comprising a natural language utterance from a user of an application associated with a set of topics and providing the natural language utterance to a natural language model configured to identify nodes of a knowledge graph. The method further includes, based on output of the natural language model, identifying a node of a knowledge graph associated with the natural language utterance, wherein the output of the natural language model includes a node identifier for the node of the knowledge graph and providing the node identifier to the knowledge engine. The method further includes receiving a response associated with the node of the knowledge graph from the knowledge engine and transmitting the response to the user in response to the long-tail query.

Abstract: Embodiments described herein provide a system for generating semantically accurate synthetic images. During operation, the system generates a first synthetic image using a first artificial intelligence (AI) model and presents the first synthetic image in a user interface. The user interface allows a user to identify image units of the first synthetic image that are semantically irregular. The system then obtains semantic information for the semantically irregular image units from the user via the user interface and generates a second synthetic image using a second AI model based on the semantic information. The second synthetic image can be an improved image compared to the first synthetic image.

Abstract: One embodiment provides a system that facilitates efficient collection of training data. During operation, the system obtains, by a recording device, a first image of a physical object in a scene which is associated with a three-dimensional (3D) world coordinate frame. The system marks, on the first image, a plurality of vertices associated with the physical object, wherein a vertex has 3D coordinates based on the 3D world coordinate frame. The system obtains a plurality of second images of the physical object in the scene while changing one or more characteristics of the scene. The system projects the marked vertices on to a respective second image to indicate a two-dimensional (2D) bounding area associated with the physical object.

Abstract: A method including receiving a user input for input to a language processing machine learning model (MLM). The method also includes generating modified inputs that are based on, and semantically related to, the user input. The method also includes executing the MLM to generate model outputs of the MLM. The MLM takes as input instances of each of the modified inputs. The method also includes sampling the model outputs using a statistical sampling strategy to generate sampled model outputs. The method also includes clustering the sampled model outputs into clusters. Each cluster of the clusters represents a distinct semantic meaning of the sampled model outputs. The method also includes generating a confidence metric for the user input. The confidence metric includes a predictive entropy of the clusters. The method also includes routing the user input based on whether the confidence metric satisfies or fails to satisfy a threshold value.