Generating or modifying text using a digital assistant and/or language model
Inventors
Edwards, Dylan R. • HANLEY, Robert W. • LAPLACA, Jennifer A.
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12468883-B2
Publication Date
2025-11-11
Expiration Date
Abstract
The present disclosure generally relates to generating and/or modifying text using a digital assistant. This disclosure encompasses a system for requesting information before generating text using models based on the type of information needed. This disclosure further encompasses a system for a remote model requesting additional information before generating text. This disclosure encompasses user interfaces for text generation/modification. This disclosure encompasses a system for proofreading text using a model. This disclosure encompasses a system for generating a response to a received communication with operations for updating the generated response.
Core Innovation
A computer system configured to communicate with a display generation component and one or more input devices receives a first input corresponding to a request for a language model to generate generated text. In response, the system determines whether additional information of a first type is to be included in the generated text, and when the additional information of the first type is to be included, the system displays a request for additional information without generating the generated text via the language model.
When the system determines that additional information of a second type is to be included in the generated text, the system generates the generated text via the language model. The generated text includes at least one placeholder associated with the additional information of the second type, and the system displays the generated text with the at least one placeholder.
After displaying the request for additional information of the first type, the system receives a second input indicating a refusal to provide the additional information of the first type. In response, the system determines whether the language model can generate the generated text, and if the language model can generate the generated text, the system generates the generated text without the additional information of the first type; if the language model cannot generate the generated text, the system displays an indication that the generated text cannot be generated.
Claims Coverage
The provided excerpts include five inventive features across independent claim sets directed to a computer system, a non-transitory computer-readable storage medium, and a method. The claims center on conditional prompting for first-type additional information, placeholder-based generation for second-type additional information, and refusal handling based on whether the language model can generate the generated text.
First-type additional information request without generating text
In response to receiving a first input requesting language-model generated text, a determination is made that additional information of a first type is to be included, and a request for the additional information is displayed without generating the generated text via the language model.
Second-type additional information using placeholders in generated text
In accordance with a determination that additional information of a second type is to be included in the generated text, generated text is generated via the language model including at least one placeholder associated with the additional information of the second type, and the generated text with the placeholder is displayed.
Refusal input triggers generation capability determination
After displaying the request for additional information of the first type, a second input is received indicating a refusal to provide the additional information of the first type.
Generation without first-type information when possible
In response to receiving the refusal input, the system determines whether the language model can generate the generated text, and when the language model can generate the generated text, the generated text is generated without the additional information of the first type.
Indication when generation cannot be performed
In response to receiving the refusal input, when the language model cannot generate the generated text, an indication is displayed via the display generation component that the generated text cannot be generated.
The claims collectively cover conditional request handling, placeholder-based text generation, and post-refusal generation capability checking. The core inventive features are the same across the independent claims.
Stated Advantages
Reduced cognitive load
Battery/processing conservation
Documented Applications
Adaptive UI expansion for text generation/proofreading based on input type and quantitative thresholds, including transitions that show expanded options
Interactive proofreading workflows correcting multiple textual errors, marking corrections, showing explanations and revert options, and allowing manual edits via reduced-clutter affordances with suggested alternative modifications
Generated text update flows and follow-on UI to revert/update generated elements
Privacy/consent and data governance including consent, anonymization/aggregation, encryption, opt-in/opt-out, and safeguards
Interested in licensing this patent?