Skip to main content
The Data Playground is an interactive exploration workspace. It loads a dataset (or subset) into an in‑memory session where you can poke at the data, run lightweight previews, and publish the result back as a new dataset or branch. It’s behind the playground.access permission, see Team & Access.

​
When to use it

  • You just imported a dataset and want a feel for it before kicking off heavy scans.
  • You need to edit labels by hand on a small subset to seed a cleaner training run.
  • You want to propose a stratified split and visualize the class balance before saving it.
  • You want a quick bias peek without paying for the full bias_shortcut engine.

​
Sessions

A session is one user’s working copy of a dataset, isolated from everyone else. Sessions live on the Playground home page (/playground).
StateWhat it means
preparingAntidote is materialising the working copy. Progress streams.
readyThe Playground editor is open and editable.
cancelledYou cancelled preparation; the session is gone.
failedPreparation errored. Retry from the home page.
idleNo edits for a while; eligible for auto‑cleanup.
Each session has its own isolated workspace so multiple teammates can explore the same dataset in parallel without stepping on each other.

​
Creating a session

1

Pick a source

From /playground click New session. Pick one of:
  • An existing dataset (whole, or filtered).
  • A new upload (zip / folder).
  • A pull from HuggingFace, Kaggle, or S3.
2

Wait for preparation

The session shows a streamed progress bar. You can cancel from the session card.
3

Open the editor

Once ready, the Playground opens with the data table, filters, and the analysis panels.

​
What the editor gives you

PaneWhat you can do
Data tablePaginate, sort, filter, inline‑edit cells. Multi‑select supports bulk label flips.
FiltersBy class, split, file size, metadata field.
Embedding viewRun an embedding pass and view a 2D projection. Lasso‑select points to filter the table.
Bias previewLightweight version of the bias engine. Directional, not authoritative.
DistributionsClass balance, file size, aspect ratio, token length, missing value counts.
SplitsPropose stratified / random / group splits and visualize the result before committing.
The bias preview and embedding view are directional, they’re meant to point you at problems, not replace a full scan. Treat them like a sanity check.

​
Saving your work

You have three ways to commit changes:
ActionWhat it does
Publish as datasetCreates a new named dataset from the working copy.
Export CSVDownloads the edited manifest. You can re‑import it later.
Update cellsCommits edits back to a source dataset’s branch.
Published datasets get a lineage edge playground_session linking them to the session they came from, so auditors can trace any hand‑edited samples back to who edited them.

​
Common workflows

  1. Create a session against the dataset, no subsetting.
  2. Open the embedding view; look for tight unexpected clusters.
  3. Skim the distributions panel for class imbalance or weird size outliers.
  4. Close the session; no need to publish.
  1. Filter the table to the class you want to clean.
  2. Use multi‑select + bulk label flip on visually mislabeled samples.
  3. Publish as a new branch on the original dataset.
  4. Train a quick model on the cured branch as an A/B test.
  1. Open the splits panel and choose stratified by class.
  2. Eyeball the proposed sizes per class.
  3. Adjust the random seed if a class is too small in a split.
  4. Update cells to commit the split column back to the dataset’s branch.

​
Sessions and cleanup

  • Long‑idle sessions time out and are cleaned up automatically. The exact threshold is set on your tenant (typically 24h of inactivity).
  • You can rename, cancel, retry, or delete any session from the Playground home page.
  • Deleting a session does not affect the dataset it was sourced from.