AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Dynamic Cluster Data Sampling for Efficient and Long-Tail-Aware Vision-Language Pre-training

Mingliang Liang, Zhuoran Liu, Arjen P. de Vries, Martha Larson

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 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 cluster scaling (up/down-sampling) to static random pruning (ablation)random pruning · ImageNet-1K zero-shot (DataComp, 0.64B seen)top-1 accuracy: 64.5 -> 68.0Improved
add cluster scaling to dynamic random sampling (ablation)Random-Dynamic · Let It Wag! zero-shot (DataComp, 0.64B seen)top-1 accuracy: 36.2 -> 46.5Improved
add dynamic per-epoch sampling to cluster scaling (ablation)Cluster-Scaling · ImageNet-1K zero-shot (DataComp, 0.64B seen)top-1 accuracy: 68.0 -> 69.2Improved
add dynamic per-epoch sampling to random pruning (ablation)random pruning · ImageNet-1K zero-shot (DataComp, 0.64B seen)top-1 accuracy: 64.5 -> 66.2Improved
add dynamic per-epoch sampling to random pruning (ablation)random pruning · Let It Wag! zero-shot (DataComp, 0.64B seen)top-1 accuracy: 35.5 -> 36.2Improved
use moderate alpha 0.2 instead of flat alpha 0.0 (ablation)DynamiCS · ImageNet-1K zero-shot (DataComp, 106M seen, 112px)top-1 accuracy: 38.5 -> 39.2Improved
raise alpha from 0.2 to 2.0 (ablation)DynamiCS · ImageNet-1K zero-shot (DataComp, 106M seen, 112px)top-1 accuracy: 39.2 -> 19.4Hurt
merge close centroids with cosine threshold 0.7 instead of none (1.0) (ablation)DynamiCS · ImageNet-1K zero-shot (DataComp, 106M seen, 112px)top-1 accuracy: 34.9 -> 39.2Improved
vary number of clusters from 10k to 70k (ablation)DynamiCS · ImageNet-1K zero-shot (DataComp, 106M seen, 112px)top-1 accuracy: 39.0 -> 39.2No clear effect
increase fine-tuning batch size from 4k to 32k (ablation)DynamiCS-1.28B · ImageNet-1K zero-shot (DataComp)top-1 accuracy: 71.1 -> 71.7Improved
add 128M samples of 224px fine-tuning after 112px pre-training (ablation)DynamiCS 1.28B (LAION-400M) · ImageNet-1K zero-shottop-1 accuracy: 60.7 -> 65.0Improved
scale samples seen from 0.64B to 2.56BDynamiCS (DataComp-DFN) · ImageNet-1K zero-shottop-1 accuracy: 69.2 -> 72.6Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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