CoolFace
Datasetpublic

disentangled-vla/libero-island-ablation-v4-basevla-aug-p05

V4 BaseVLA + Domain Augmentation (p=0.5) — LIBERO Island Eval Ablation V4 of the Stage3 Qwen2.5-VL framework: BaseVLA with image-level domain augmentation (lighting + warp randomization, applied with probability p=0.5). Evaluated on the LIBERO-Spatial island viewpoint protocol from Xing et al. 2025 (CoRL) "Shortcut Learning in Generalist Robot Policies". Architecture Backbone: Qwen/Qwen2.5-VL-3B-Instruct + LoRA (r=32, q/v_proj) Single AttentiveLatentHead A (8… See the full description on the dataset page: https://huggingface.co/datasets/disentangled-vla/libero-island-ablation-v4-basevla-aug-p05.

sourceHugging Faceapache-2.0updated 4mo agoView on Hugging Face
0likes30downloads
root
dirvideos/
file.gitattributes2 KBdownload
fileAA_v0.200-0.350_tasks0,1,3,5,8_results.jsonl6 KBdownload
fileAA_v0.200-0.350_tasks0,1,3,5,8_summary.json797 Bdownload
fileAL_B_v0.200-0.200_tasks2,4,6,7,9_results.jsonl6 KBdownload
fileAL_B_v0.200-0.200_tasks2,4,6,7,9_summary.json795 Bdownload
fileBB_v0.650-0.800_tasks2,4,6,7,9_results.jsonl6 KBdownload
fileBB_v0.650-0.800_tasks2,4,6,7,9_summary.json796 Bdownload
fileBR_A_v0.800-0.800_tasks0,1,3,5,8_results.jsonl6 KBdownload
fileBR_A_v0.800-0.800_tasks0,1,3,5,8_summary.json796 Bdownload
fileC_A_v0.500-0.500_tasks0,1,3,5,8_results.jsonl6 KBdownload
fileC_A_v0.500-0.500_tasks0,1,3,5,8_summary.json796 Bdownload
fileC_B_v0.500-0.500_tasks2,4,6,7,9_results.jsonl6 KBdownload
fileC_B_v0.500-0.500_tasks2,4,6,7,9_summary.json796 Bdownload
fileREADME.md2 KBdownload

disentangled-vla/libero-island-ablation-v4-basevla-aug-p05 · main · files are served by the source, never re-hosted here