DP-FM decouples radial and angular dynamics on a cylindrical manifold via constant-warping metric and classifier-free guidance to achieve state-of-the-art multi-step few-shot adaptation of vision-language models on 11 benchmarks.
Classifier-free diffusion guidance
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
StitchVM stitches clean-image reward models with diffusion backbones to enable efficient value estimation for noisy latents, speeding up diffusion alignment methods like DPS by 3.2x and halving memory.
citing papers explorer
-
Direct Product Flow Matching: Decoupling Radial and Angular Dynamics for Few-Shot Adaptation
DP-FM decouples radial and angular dynamics on a cylindrical manifold via constant-warping metric and classifier-free guidance to achieve state-of-the-art multi-step few-shot adaptation of vision-language models on 11 benchmarks.
-
Stitched Value Model for Diffusion Alignment
StitchVM stitches clean-image reward models with diffusion backbones to enable efficient value estimation for noisy latents, speeding up diffusion alignment methods like DPS by 3.2x and halving memory.