CoolFace
Datasetpublic

bag100/action-atlas-xvla

Action Atlas: X-VLA sparse autoencoders and concepts Sparse autoencoders (SAEs) and identified concepts for X-VLA, part of the Action Atlas release accompanying the paper on cross-task activation injection in vision-language-action models. The interactive explorer is at https://action-atlas.com. What is here saes/ TopK SAEs (k=64, 8x expansion) over the X-VLA (soft-prompted Florence-Large), 24 layers, 1024-dim. Arms present: per-token 48, mean-pool 48. Per-token… See the full description on the dataset page: https://huggingface.co/datasets/bag100/action-atlas-xvla.

sourceHugging Facecc-by-4.0updated 3mo agoView on Hugging Face
0likes235downloads

bag100/action-atlas-xvla · main · files are served by the source, never re-hosted here