SOURCE-LINKED INTELLIGENCE

Dreaming in Flow: Generative Grounding Feedback for Self-Evolving Unified Multimodal Models

arXiv · AI, language, vision and robotics · article · Sep 8, 2026 · UTC

Unified multimodal models integrate visual understanding and generation within a single network, yet the two capabilities are commonly optimized as separate tasks. We introduce Generative Grounding Feedback(GGF), a self-evolving post-training framework that uses only text prompts and the model's own visual experience. Given a prompt, the model first generates a visual ``dream.'' Flow-level feedback compares text-, image-, and repair-conditioned predictions at the same noisy latent state, transferring image-grounded generation directions to the prompt condition. Dream replay grounding replays t

Read original source ↗ Open in workspace

recordType

paper

region

Global