AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Dynamic Classifier-Free Diffusion Guidance via Online Feedback

Pinelopi Papalampidi, Olivia Wiles, Ira Ktena, Aleksandar Shtedritski, Emanuele Bugliarello, Ivana Kajic, Isabela Albuquerque, Aida Nematzadeh

ICLR 2026 · Image generation › Text-to-image diffusion · approach: Inference-time T2I improvement · 12 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 adaptive timestep-dependent weighting instead of linear weighting of alignment and VQ evaluators (ablation)Dynamic CFG search (LDM-large) · Gecko promptsGecko score: 45.0 -> 47.2Improved
use adaptive weighting instead of linear weighting (ablation)Dynamic CFG search (LDM-large) · MS COCO promptsFID: 25.4 -> 24.8Improved
replace per-prompt dynamic schedule with static mean schedule of dynamic CFG (ablation)Dynamic CFG search (LDM-large) · Gecko promptsGecko score: 47.2 -> 46.5Hurt
replace per-prompt dynamic schedule with static mean schedule (ablation)Dynamic CFG search (LDM-large) · MS COCO promptsFID: 24.8 -> 26.8Hurt
add dynamic CFG search with adaptive Alignment + VQ evaluators to default fixed CFGLDM-large · Gecko promptsGecko score: 43.8 -> 47.2Improved
add dynamic CFG search with adaptive Alignment + VQ evaluators to default fixed CFGLDM-large · MS COCO promptsFID: 25.6 -> 24.8Improved
use dynamic CFG with only the alignment evaluator vs only the VQ evaluator (ablation)Dynamic CFG search (LDM-large) · Gecko prompts and MS COCOGecko score / FID: 45.5 / 26.4 (alignment) vs 44.0 / 24.8 (VQ)Mixed
add gradient guidance combining alignment and VQ models to default CFGLDM-large · Gecko promptsGecko score: 43.8 -> 45.3Improved
apply annealing CFG schedule instead of fixed CFGLDM-large · MS COCO promptsFID: 25.6 -> 28.9Hurt
filter with pixel-space CLIP instead of latent CLIP at 50% of denoising (ablation)LDM-small · Gecko promptsGecko score: 41.4 -> 44.6Improved
filter samples (best of 4) with latent VQ discriminator at 25% of denoisingLDM-large · MS COCOFID: 29.2 -> 27.6Improved
add alignment and reward evaluators with adaptive weighting to Imagen 3 samplingImagen 3 · Geckohuman win rate (%): 53.6Improved

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