turhancan97/SpaRRTa
SpaRRTa: A Synthetic Benchmark for Evaluating Spatial Intelligence in Visual Foundation Models SpaRRTa is a synthetic benchmark that probes whether Visual Foundation Models (VFMs) — such as DINO, DINOv2/v3, MAE, CroCo, VGGT, SPA and CLIP — encode the spatial relations between objects in a scene, rather than only their semantic identity. 📄 Paper: arXiv:2601.11729 💻 Code: github.com/gmum/SpaRRTa 🧱 Real-world (lego) split: turhancan97/SpaRRTa-Lego 🔬 Attention-analysis split (images +… See the full description on the dataset page: https://huggingface.co/datasets/turhancan97/SpaRRTa.
Update README.md
Update README.md
Update README.md
Upload teaser and pipeline
Update README.md
Update README.md
Update README.md
Update README.md
Update README.md
Update README.md
Update README.md
Update README.md
Update README.md
Upload logo.png
Update README.md
Add 9881 samples in 39 parquet shard(s) (2026-03-10 00:29:07 UTC)
Add 9881 samples in 39 parquet shard(s) (2026-03-10 00:24:11 UTC)
Add 9881 samples in 39 parquet shard(s) (2026-03-09 23:56:50 UTC)
Add 20000 samples in 79 parquet shard(s) (2026-03-09 23:51:01 UTC)
Add 30000 samples in 118 parquet shard(s) (2026-03-09 22:41:43 UTC)
Add 30000 samples in 118 parquet shard(s) (2026-03-09 22:13:07 UTC)
Add 10000 samples in 40 parquet shard(s) (2026-03-09 21:56:43 UTC)
Add 29502 samples in 116 parquet shard(s) (2026-03-09 21:45:56 UTC)
initial commit