Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance

2026-06-25

AI Summary

This work addresses the issue of diversity collapse in pre-trained flow models during batch conditional generation. To mitigate this, the authors propose a training-free self-guidance mechanism that enhances sample diversity by dispersing internal features of the flow model during inference, while simultaneously employing manifold regularization to project these perturbed features back onto the data manifold, thereby preserving conditional alignment. The method is plug-and-play, requires no external reward models, and incurs minimal inference overhead. It achieves, for the first time, efficient diversity enhancement under multimodal conditions—including text prompts, depth maps, and reference images—without compromising fidelity. Experimental results demonstrate significant improvements in generation diversity across both multi-step and few-step sampling regimes, while maintaining high sample quality.

Abstract

State-of-the-art flow models generate stunning images from text or image prompts. However, they suffer from diversity collapse when generating multiple samples under the same conditioning. Existing methods address this issue via either latent guidance, which has limited effectiveness, or sample selection, which relies on external reward models that incur significant inference-time overhead. In this work, we introduce an efficient, training-free self-guidance mechanism to mitigate diversity collapse without requiring additional reward models. Specifically, we disperse the internal features of the flow model during batch generation with feature self-guidance. Further, to keep the features close to the manifold, we introduce a manifold regularization step that projects these dispersed features back onto the data manifold, ensuring diverse generation without sacrificing alignment with the input conditions. Our method integrates seamlessly as a plug-and-play module into pretrained flow models, adding only a marginal inference cost. Experiments demonstrate significant improvements in diversity while preserving fidelity across several conditional flow models, including multi-step and few-step text-to-image, depth-to-image, and reference image generation.

Problem

• diversity collapse
• flow models
• conditional generation
• sample diversity
• pretrained models

Innovation

• feature self-guidance
• diversity collapse
• flow models
• manifold regularization
• training-free