Systems and methods for modification of machine learning model-generated text and images based on user queries and profiles

A method comprising:

• providing user data associated with a user and a first query to one or more machine learning models to obtain a plurality of outputs and a corresponding plurality of images;
• generating a plurality of graphical representations for the plurality of outputs and the corresponding plurality of images;
• obtaining a selection of a first graphical representation of the plurality of graphical representations;
• determining a plurality of requested characteristics describing a requested item associated with the first graphical representation based on the first graphical representation
• comparing the plurality of requested characteristics with a plurality of characteristic sets to determine a plurality of items, wherein each set comprises a respective item characteristic associated with a corresponding item;
• generating a plurality of representative descriptions corresponding to the plurality of items;
• and
• based on receiving a selection of a first representative description of the plurality of representative descriptions, enabling access to a first item corresponding to the first representative description.

One or more non-transitory, computer-readable media storing instructions that, when executed by one or more processors, cause operations comprising:

• obtaining, based on a query, an item description of a first item;
• providing the query and the item description to one or more machine-learning models to cause the one or more machine-learning models to generate output comprising at least one of a first text output or a first synthetic image associated with the first item;
• determining whether one or more generated characteristics of the output satisfy a set of similarity criteria associated with the first item;
• based on determining that the one or more generated characteristics do not satisfy the set of similarity criteria, constructing corrected output for the first item by modifying or regenerating at least a portion of the output based on the set of similarity criteria;
• and
• updating the one or more machine-learning models based on the corrected output to reduce likelihood of generating later output that does not satisfy similarity criteria associated with an item description corresponding to a later query.