AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Efficient Tail-Aware Generative Optimization via Flow Model Fine-Tuning

Zifan Wang, Riccardo De Santi, Xiaoyu Mo, Michael M. Zavlanos, Andreas Krause, Karl Henrik Johansson

ICML 2026 · Image generation › Fast diffusion sampling · approach: Rectified flow straightening · 0 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
Use left-CVaR pseudo-reward (L-TFFT) instead of expected-reward fine-tuningStable Diffusion v1-5 · ImageRewardL-CVaR0.2[r]: -0.370 -> -0.306Improved
Use L-TFFT instead of EXP-FTStable Diffusion v1-5 · ImageRewardE[r]: 0.852 -> 0.850No clear effect
Use L-TFFT instead of FDC under fixed 82 GPU hoursStable Diffusion v1-5 · ImageRewardE[r]: 0.812 -> 0.850Improved
Use L-TFFT instead of EXP-FTStable Diffusion v1-5 · DreamSim diversityDreamSim Var: 0.306 -> 0.289Hurt
Use L-TFFT instead of EXP-FTStable Diffusion v1-5 · HPSv2HPS: 0.278 -> 0.280Improved
Use right-CVaR (R-TFFT) instead of EXP-FTFlowMol · xTB energy rewardR-CVaR0.9[r]: 167.1 -> 183.4Improved
Use R-TFFT instead of FDC (K=3)FlowMol · xTB energy rewardR-CVaR0.9[r]: 172.5 -> 183.4Improved
Use R-TFFT instead of EXP-FTFlowMol · xTB energy rewardE[r]: 29.8 -> 27.1Hurt
Use R-TFFT instead of EXP-FTFlowMol · xTB energy rewardValidity: 78.3%±1.01% -> 85.7%±0.59%Improved
Use R-TFFT instead of EXP-FTFlowMol · xTB energy rewardSA score: 7.66 -> 7.70No clear effect
Solve CVaR dual threshold once offline instead of iterative FDCTFFT · training costfine-tuning calls: K -> 1Improved

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