AI Paper Atlas

Research map / Large language models / Alignment and preferences

K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge

Zhikai Li, jiatong li, Xuewen Liu, Wangbo Zhao, Pan Du, Kaicheng Zhou, Qingyi Gu, Yang You, Zhen Dong, Kurt Keutzer

ICLR 2026 · Large language models › Alignment and preferences · approach: LLM-as-judge and feedback data · 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
add posterior correction to the VLM judge (ablation)K-Sort Eval · Ablation T2I (FLUX.1-dev)score: 29.32 -> 28.86Improved
add posterior correction to the VLM judge (ablation)K-Sort Eval · Ablation T2V (CogVideoX-5b)score: 31.86 -> 33.63Improved
add dynamic matching (ablation)K-Sort Eval · Ablation T2I (FLUX.1-dev)runs: 500 -> 81Improved
add dynamic matching (ablation)K-Sort Eval · Ablation T2V (CogVideoX-5b)runs: 300 -> 89Improved
add dynamic matching (ablation)K-Sort Eval · Ablation T2I (FLUX.1-dev)score: 28.79 -> 28.86No clear effect
add wrapping operation to the prompt (ablation)K-Sort Eval · Ablation T2I (FLUX.1-dev)score: 28.93 -> 28.86Improved
add wrapping operation to the prompt (ablation)K-Sort Eval · Ablation T2V (CogVideoX-5b)score: 33.55 -> 33.63No clear effect
add rule augmentation to the prompt (ablation)K-Sort Eval · Ablation T2I (FLUX.1-dev)score: 28.13 -> 28.86Improved
add rule augmentation to the prompt (ablation)K-Sort Eval · Ablation T2V (CogVideoX-5b)score: 33.10 -> 33.63Improved
add posterior correction to GPT-4o judge when correlating with human preferenceGPT-4o VLM judge · 100 instance groupscorrelation with human preferencesImproved
use GPT-4o judge instead of CLIP-based scorerspreference scoring · 100 instance groupscorrelation with human preferencesImproved
distill SD-v3.5-large to turbo (4 steps)SD-v3.5-large · Compressed modelsscore: 28.95 -> 27.71Hurt

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