Navigating User Behavior toward Personalized Multimodal Generation
This work addresses the misalignment between user intent and AIGC outputs for end users who cannot articulate visual details, enabling more personalized multimodal generation.
NaviGen improves personalized image and video generation by encoding user behavior into a dual identifier (collaborative + textual code) and using a two-stage SFT+RL pipeline, achieving better next-item prediction and more specific, generatable instructions across product, game, and short-video domains.
Modern AIGC pipelines deliver high-fidelity images and videos but presuppose a well-formed creation instruction, while end users rarely articulate visual details, leaving generators misaligned with user demand. We study personalized content generation, which turns a user's interaction history into an executable instruction for downstream synthesis, and identify two obstacles: behavior must be encoded in a form legible to language reasoning, and the model must acquire instruction-writing skill absent from both pretraining and behavior data. We propose NaviGen, which represents each item with a dual identifier coupling a collaborative code and a textual code as a behavioral substrate and a semantic bridge in one token stream. On this representation, a two-stage SFT+RL pipeline first distills preference reasoning and instruction writing from evolutionarily searched supervision, then aligns generation with user intent through hierarchical and self-consistent rewards. Experiments across product, game, and short-video domains show that NaviGen improves personalized image and video generation, strengthens next-item prediction, and yields more specific, relevant, and visually generatable instructions. Our code is anonymously released at: https://github.com/iLearn-Lab/NaviGen.