DanceOPD: On-Policy Generative Field Distillation

Hugging Face Daily Papers Papers

Summary

DanceOPD proposes an on-policy generative field distillation framework for flow-matching models that unifies text-to-image generation, local editing, and global editing via capability-specific routing and velocity-based training, improving multi-capability composition while preserving anchor generation quality.

Modern image generation demands a single model that unifies diverse capabilities, including text-to-image (T2I), local editing, and global editing. However, these capabilities are rarely naturally aligned and often conflict. For instance, editing tends to degrade T2I performance, while global and local editing interfere with each other. Consequently, effectively composing these capabilities has become a central challenge for image generation model training. To tackle this, we introduce DanceOPD, an on-policy generative field distillation framework for flow-matching models that routes each sample to one capability field, queries one low-noise student-induced state, and trains with a simple velocity MSE objective. With each capability source defined as a velocity field over the shared flow state space, the student learns from fields queried on its own rollout states to compose expert capabilities. This formulation also absorbs operator-defined fields such as classifier-free guidance. Comprehensive experiments on T2I, editing, realism-field absorption, and CFG absorption show that our approach improves multi-capability composition, strengthening target capabilities while preserving anchor generation quality. We believe this work establishes a practical route for generative field distillation in flow-matching models.
Original Article
View Cached Full Text

Cached at: 06/26/26, 02:04 AM

Paper page - DanceOPD: On-Policy Generative Field Distillation

Source: https://huggingface.co/papers/2606.27377

Abstract

A novel on-policy generative field distillation framework called DanceOPD is proposed to unify text-to-image generation, local editing, and global editing capabilities in flow-matching models through capability-specific routing and velocity-based training.

Modern image generation demands a single model that unifies diverse capabilities, includingtext-to-image(T2I),local editing, andglobal editing. However, these capabilities are rarely naturally aligned and often conflict. For instance, editing tends to degrade T2I performance, while global andlocal editinginterfere with each other. Consequently, effectively composing these capabilities has become a central challenge for image generation model training. To tackle this, we introduce DanceOPD, anon-policygenerative field distillationframework forflow-matching modelsthat routes each sample to one capability field, queries one low-noise student-induced state, and trains with a simplevelocity MSE objective. With each capability source defined as avelocity fieldover the shared flow state space, the student learns from fields queried on its own rollout states to composeexpert capabilities. This formulation also absorbs operator-defined fields such asclassifier-free guidance. Comprehensive experiments on T2I, editing, realism-field absorption, and CFG absorption show that our approach improves multi-capability composition, strengthening target capabilities while preserving anchor generation quality. We believe this work establishes a practical route forgenerative field distillationinflow-matching models.

View arXiv pageView PDFProject pageAdd to collection

Get this paper in your agent:

hf papers read 2606\.27377

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.27377 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.27377 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.27377 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Flow-OPD: On-Policy Distillation for Flow Matching Models

Hugging Face Daily Papers

Flow-OPD is a research paper introducing a two-stage on-policy distillation framework for Flow Matching text-to-image models, significantly improving generation quality and alignment metrics using Stable Diffusion 3.5 Medium.

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation

Hugging Face Daily Papers

AnyFlow introduces a novel any-step video diffusion distillation framework that optimizes full ODE sampling trajectories through flow-map transition learning and backward simulation, achieving performance that matches or surpasses consistency-based counterparts while scaling with sampling step budgets.

Draft-OPD: On-Policy Distillation for Speculative Draft Models

Hugging Face Daily Papers

Draft-OPD introduces on-policy distillation with target-assisted rollouts and error replay to overcome the offline-to-inference mismatch in training draft models for speculative decoding, achieving over 5x lossless acceleration and improving upon EAGLE-3 and DFlash by 23% and 13% respectively.