AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

ShortFT: Diffusion Model Alignment via Shortcut-based Fine-Tuning

Xiefan Guo, Miaomiao Cui, Liefeng Bo, Di Huang

ICCV 2025 · Reinforcement learning › RL from feedback · approach: RL fine-tuning of generative models · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add timestep-aware LoRA (ablation)ShortFT · HPDv2HPS v2: 33.46 -> 33.88Improved
add timestep-aware LoRA (ablation)ShortFT · HPDv2PickScore: 23.82 -> 24.16Improved
add timestep-aware LoRA (ablation)ShortFT · HPDv2Symmetry: 0.187 -> 0.138Improved
add progressive training strategy (ablation)ShortFT · HPDv2HPS v2: 33.27 -> 33.88Improved
add progressive training strategy (ablation)ShortFT · HPDv2PickScore: 23.97 -> 24.16Improved
add progressive training strategy (ablation)ShortFT · HPDv2Symmetry: 0.146 -> 0.138Improved
fine-tune foundational SD 1.5 instead of few-step Hyper-SD (ablation)ShortFT · HPDv2HPS v2: 32.92 -> 35.97Improved
use ShortFT instead of DRTune under equal computebackpropagation fine-tuning · HPDv2HPS v2: 32.79 -> 33.88Improved
use ShortFT instead of DRaFT-LV under equal computebackpropagation fine-tuning · HPDv2Symmetry: 0.418 -> 0.138Improved
use ShortFT instead of DRTune under equal computebackpropagation fine-tuning · HPDv2Symmetry: 0.207 -> 0.138Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy