Link the MulTaBench paper and leaderboard, and add the citation

README.md

@@ -0,0 +1,30 @@
+---
+tags:
+- multabench
+- embeddings
+---
+
+# MulTaBench text embeddings (e5-small)
+
+Part of [MulTaBench](https://github.com/alanarazi7/MulTaBench), a benchmark for multimodal tabular learning.
+
+Frozen text embeddings of the MulTaBench datasets, cached so leaderboard runs don't re-encode them. Each
+`<dataset>/embeddings.npz` maps every distinct input, exactly as the encoder got it, to its raw encoder output (PCA is
+fitted per run, on the training rows); `meta.json` holds the dataset's rows and columns, the encoding time per feature
+and the hardware. See [Cached embeddings](https://github.com/alanarazi7/MulTaBench#cached-embeddings).
+
+**Paper:** [MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image](https://huggingface.co/papers/2605.10616) · **Leaderboard:** [multabench/leaderboard](https://huggingface.co/spaces/multabench/leaderboard)
+
+## Citation
+
+```bibtex
+@misc{arazi2026multabenchbenchmarkingmultimodaltabular,
+  title={MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image},
+  author={Alan Arazi and Eilam Shapira and Shoham Grunblat and Mor Ventura and Elad Hoffer and Gioia Blayer and David Holzmüller and Lennart Purucker and Gaël Varoquaux and Frank Hutter and Roi Reichart},
+  year={2026},
+  eprint={2605.10616},
+  archivePrefix={arXiv},
+  primaryClass={cs.LG},
+  url={https://arxiv.org/abs/2605.10616},
+}
+```