AI Paper Atlas

Research map / Benchmarks / VTAB-1k Specialized

VTAB-1k Specialized leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on VTAB-1k Specialized. Best accuracy: 90.25 (LSPT + pMoE, ICLR 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

VTAB-1k Specialized · accuracy

10 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
190.25LSPT + pMoE
ImageNet-21K ViT-B/16
pMoE: Prompting Diverse Experts Together Wins More in Visual Adaptation
ICLR 2025 · Table: VTAB-1K supervised
Pruning and efficient tuning
286.4MIST
ViT-B/16
Minimal Interaction Seperated Tuning: A New Paradigm for Visual Adaptation
CVPR 2025 · Table 1
Pruning and efficient tuning
386.3SA2VP
Swin Transformer
SA²VP: Spatially Aligned-and-Adapted Visual Prompt
AAAI 2024 · Table 4
CLIP and zero-shot
486.15VFPT
Swin-Base, supervised ImageNet-21k
Visual Fourier Prompt Tuning
NeurIPS 2024 · Table 2
CLIP and zero-shot
585.7LIM + LoRA + membrane transfer (full)
Vim-S
Memba: Membrane-driven Parameter-Efficient Fine-Tuning for Mamba
ICLR 2026 · Table 3(a)
Pruning and efficient tuning
684IN21k + JFT All Experts
ResNet50-v2
Scalable Transfer Learning with Expert Models
ICLR 2021 · VTAB-1k results table
Recognition and long-tail
783.78DAM-VP
ViT-B/16
Diversity-Aware Meta Visual Prompting
CVPR 2023 · Table: VTAB benchmark
Zero-/few-shot and meta-learning
883.33VPT
Swin-Base, ImageNet-21k
Facing the Elephant in the Room: Visual Prompt Tuning or Full finetuning?
ICLR 2024 · Swin-Base Specialized table
CLIP and zero-shot
982.96Hybrid (w/ proj)
Vim-S
MambaPEFT: Exploring Parameter-Efficient Fine-Tuning for Mamba
ICLR 2025 · Table 1
Pruning and efficient tuning
1082.28Deep-share_DEEP
ViT-B/16
Revisiting Prefix-tuning: Statistical Benefits of Reparameterization among Prompts
ICLR 2025 · Table: FGVC and VTAB-1K result
Pruning and efficient tuning

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.