Capital One

Systems and methods for improving component access for cloud service users. The systems and methods allow for users to access additional cloud components through data layer operations such as routing table updates. Notably, the data layer may be modified by the cloud service users themselves and thus avoid restriction in inactivity by the cloud service provider.

A method and related system may establish communication guardrails during multi-turn conversations. The method may include constructing a graph indicating a set of intents and a set of intent states by inputting a set of transcripts to a language model and determining a set of metrics of events by traversing a path through nodes of the graph based on intent states indicated by the graph.

A device may obtain a plurality of responses of a language model to a prompt, where each of the plurality of responses comprises a binary response portion and a reasoning portion. The device may determine an uncertainty quantification for the language model based on the plurality of responses. The uncertainty quantification may be based on variations across reasoning portions of the plurality of responses, or variations in binary option confidences in binary response portions of the plurality of responses. The device may output the binary response portion of one of the plurality of responses based on the uncertainty quantification indicating certainty in the plurality of responses, or an uncertainty indication based on the uncertainty quantification indicating uncertainty in the plurality of responses.

Methods and systems are described herein for managing prompts used for executing and evaluating supervised fine tuning (SFT) processes. For example, a model identifier and a first description of a first computing task may be extracted from a first request to execute an SFT process to perform the first computing task. A candidate prompt template and associated metadata including a candidate prompt name may be generated, and the candidate prompt template may be uploaded to a prompt database with the associated metadata. A second request to execute an SFT process indicating the candidate prompt name may be received. The candidate prompt template may be used to execute the SFT process on a large language model associated with the second request, resulting in a trained first large language model. The trained first large language model may be evaluated using a test data set and the candidate prompt template.

Systems and methods for generating plain text descriptions by translating predictions based on speech signal or data processing from similar users. The system may retrieve a first embedding set, wherein the first embedding set comprises a plurality of user embeddings, wherein each user embedding of the plurality of user embeddings, correspond to respective user query of a plurality of user queries. The system may process the plurality of user embeddings with a first classifier model to generate a first subset of user embeddings corresponding to a first prediction of a prediction pairing and a second subset of user embeddings corresponding to a second prediction of the prediction pairing. The system may determine, using the first classifier model, that a first user query corresponds to the first prediction. The system may determine a first response to the first user query based on the prediction pairing.

Methods and systems are described herein for facilitating queryless information retrieval during a live conversation via a bifurcated model. For example, the system may receive, a first set of decrypted utterances spoken during the live conversation. The system may provide the first set of decrypted utterances as input to a bifurcated model comprising (i) a first bifurcated model portion associated with a dynamic utterance embedding space and (ii) a second bifurcated model portion associated with a static computer file embedding space. The system may then generate, via the bifurcated model, a first intermediate output from the first bifurcated model portion indicating a first utterance embedding and a second intermediate output from the second bifurcated model portion indicating a first computer file embedding. The system may then generate during the live conversation, a graphical representation of a computer file associated with the first computer file embedding.

Described herein are systems, methods, and programming for determining a context source of a large language model. For example, the techniques leverage a token accessible large language model to extract token activations associated with answers generated by the model in response to prompts. The token activations can be passed to a trained classification model that learns patterns of token activations associated with context-based model outputs. The trained classification model can assign a label to the model-produced answer indicating whether the answer was generated by the large language model using the input context included by the prompt or model memory of the large language model.

A device may receive an input indicating a prompt for a language model. The device may determine, using the language model, an output responsive to the prompt. The device may determine a degree by which the output relies on memory of the language model. The device may output, responsive to the degree by which the output relies on memory indicating an over-reliance on memory, a modified output that differs from the output.

A method and related system may reduce hallucinatory text generation by using shuffled training data.

In some implementations, a system may obtain, based on a graphical user interface (GUI) that includes a first field identifying a plurality of transformation objects, first input that indicates selection of a transformation object of the plurality of transformation objects. The system may obtain, based on a set of one or more second fields of the GUI that are associated with identifying parameter information associated with the transformation object, second input that indicates one or more parameters associated with the transformation object. The system may generate a transformation data structure, to be processed in a transformation sequence, that indicates the transformation object and the one or more parameters associated with the transformation object.