chk
Datasets
All datasets matching “chk”stvsr-chki50-archive
stvsr-chki50-archive
CHKI 50-clip full-clip experiment: prediction PNG tars and all experiment results (tcl2).
Status: placeholder (2026-09-15). Content is being added in verified batches.
Layout (planned)
MANIFEST.jsonl - one JSON line per archived item: local_path, repo, path_in_repo, bytes, sha256, n_files, encrypted, verified_remote, verified_download, deleted_utc.
Tar archives of PNG trees (per-file md5 lists kept in the manifest side files).
*.tar.enc -… See the full description on the dataset page: https://huggingface.co/datasets/MingzhenL/stvsr-chki50-archive.diffusion_policy_robocasa_activations_latest_chkpt
Diffusion Policy — RoboCasa Activations (latest checkpoint)
Per-step, per-episode activation traces collected from a DiffusionTransformerHybridImagePolicy (the diffusion_policy library) rolled out on RoboCasa benchmark tasks. Captured with collect_activations_robocasa.py at the latest training checkpoint.
These traces are the input expected by the conceptor / SAE steering pipelines under diffusion_policy/experiments/robocasa_steering/ and diffusion_policy/experiments/sae/ — see… See the full description on the dataset page: https://huggingface.co/datasets/brandonyang/diffusion_policy_robocasa_activations_latest_chkpt.chki50-h266ip-questeval_so101_006_chk085000This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 50,
"total_frames": 83624,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:50"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/kartoffelchips/eval_so101_006_chk085000.gs_physics
GS Physics Dataset
This dataset accompanies the CVPR 2026 Findings paper
Learning a Particle Dynamics Model with Real-world Videos.
It contains real-world multi-object interaction recordings and processed
Gaussian Splatting particle-dynamics data for the bowling and cube_stacks
scenarios.
The release includes two archives:
dataset.z01 and dataset.zip: two volumes of the raw dataset archive, containing multi-view RGB/stereo/depth recordings,
segmentation masks, camera metadata… See the full description on the dataset page: https://huggingface.co/datasets/chkim403/gs_physics.psm-memory-training-data
PSM Memory — training data & eval fixtures
Training curricula and evaluation fixtures used to train the storage/retrieval-plan/consolidation
LoRA adapters at chkrishna2001/psm-memory-qwen0.5b.
training-data/
All training-round JSONL curricula across every iteration of the three adapters (storage v1-v16b,
consolidation v1-v9, retrieval-plan v1-v3), plus raw source conversation exports (ChatGPT/Codex/Gemini
transcripts mined for training examples) and… See the full description on the dataset page: https://huggingface.co/datasets/chkrishna2001/psm-memory-training-data.