Cached text and image embeddings of the datasets, one repo per encoder; the -tar ones per fold.
AI & ML interests
None defined yet.
Recent Activity
View all activity
Cached text and image embeddings of the datasets, one repo per encoder; the -tar ones per fold.
The 80 MulTaBench datasets: tables with text and image columns, core and extended.
Benchmarking multimodal tabular learning with text and images (NeurIPS 2026 Spotlight): leaderboard, 80 datasets and cached embeddings.