Rethinking Classifier-Free Guidance in On-Policy Diffusion Distillation
- Published
- Source
- arXiv
- Paper number
- 730
- Field
- Computer Vision
- arXiv ID
- 2607.24731
Key points
- OPD, which only matches CFG synthesis, fails to capture branch-level stagnation, so errors cancel out.
- Privileged information in the teacher's negative branch causes the negative-branch asymmetry, or NBA, failure.
- PDM separately constrains the positive prediction and the CFG direction, meaning positive minus negative, to prevent cancellation.
- When the guidance scale changes, the baseline method breaks, but PDM remains stable.
- By adjusting direction weights and the number of supervised states, it finds the best trade-off between accuracy and efficiency.
- It improves control fidelity in pose, depth, and dense versus sparse scribble video control.
Paper links
External research summaries. These are not HDATF publications or measured product results.