Kkuntal990/vbvr-thinkgen

VBVR ThinkGen

VBVR ThinkGen is the current ThinkGen training snapshot derived from the balanced VBVR corpus. It contains 99,958 unique videos across 100 generator families, paired with the v7 prompts used by the ThinkGen video-reasoning experiments.

The snapshot combines:

• —the prompt revisions from `May-apple/VBVR-Reorganized`,
• —the additive ThinkGen v3-v5 prompt revisions over previously untouched families, and
• —the v6 owner-review pass (2026-09-08): 30 families / 30,000 prompts edited on top of v5 — 14 fixes to already-revised families (restoring task-process clauses such as "show the full process step by step", "following the law of reflection", "circle ... with a red circle" that earlier passes had deleted, plus further trims) and 16 first-time trims of families no earlier pass had touched (removing scene details readable from the first frame: object colours/shapes, chart titles and axis labels, the current clock time, direction lists, explanatory glosses).
• —the v7 checklist desk pass (2026-09-10): 21 families / 20,380 prompts edited on top of v6 after applying the reasoning-inducing prompt checklist to all 100 families — 5 defect repairs (G-45 stray periods in every row, O-7 "a unknown" colour tokens in 998 rows, O-8 wrong answer-slot wording, G-27 stranded articles, O-1 process clause), 11 restorations of task-process, invariant and stopping-rule clauses that the Reorganized pass had deleted (G-46, G-133, G-134, O-4, O-13, O-14, O-21, O-23, O-24, O-29, O-31), 2 well-posedness additions (G-12 leg convention verified in the generator source; G-51 "The colors repeat in a fixed cycle."), and 3 leak/echo removals (O-12 transformation narrative and opaque colour ids, O-75 numeric level list, O-45 rule-type name).

The video set is matched to the balanced ThinkGen training subset; prompt revisions do not replace or regenerate the source videos.

Versions

Every prompt snapshot is a git tag on this repo; the archives never change between versions.

To train on the previous prompts, fetch the metadata at the tag:

hf download Kkuntal990/vbvr-thinkgen metadata.csv --repo-type dataset --revision v6

The v6 and v7 edits are fully specified as per-family regex rule files (scripts/data/review_edits_v6/, scripts/data/review_edits_v7/) and applied by scripts/data/apply_review_edits_v6.py in the Kkuntal990/vbvr-realignment repository; re-running them on the v5 and v6 metadata reproduces each file byte-for-byte.

Files and layout

metadata.csv contains:

Prompts can contain embedded newlines. Read metadata.csv with an RFC-compliant CSV parser rather than line-oriented shell tools.

The media are stored as 100 uncompressed family archives under archives/. MP4 is already compressed, so uncompressed tar avoids wasted CPU and permits fast extraction. Each archive preserves the paths referenced by metadata.csv, for example:

G-11_handle_object_reappearance_data-generator/
  handle_object_reappearance_task/
    handle_object_reappearance_00000000/
      ground_truth.mp4

After downloading a snapshot, extract the archives into the snapshot root:

for archive in archives/*.tar; do tar -xf "$archive"; done

The video values in metadata.csv then resolve directly from that root.

Snapshot statistics

• —Rows: 99,958
• —Unique videos: 99,958
• —Generator families: 100
• —Video payload: 30,467,788,636 bytes
• —Prompt revision: ThinkGen v7
• —Families whose prompts differ from the original VBVR text: 70 (28 upstream-reorganized, 42 additive ThinkGen)
• —Families edited in v6 relative to v5: 30 (30,000 prompts)
• —Families edited in v7 relative to v6: 21 (20,380 prompts)

dataset_manifest.json records the metadata hash, snapshot-level counts and the previous snapshot's hash/tag. archives_manifest.json records the byte size, SHA-256, and video count for every family archive.

Loading metadata

from datasets import load_dataset

ds = load_dataset("Kkuntal990/vbvr-thinkgen", data_files="metadata.csv", split="train")

Use snapshot_download or hf download to fetch the complete media snapshot before extraction.

Provenance and license

The source corpus is `Video-Reason/VBVR-Dataset`. Reorganized prompts come from `May-apple/VBVR-Reorganized`. Both source repositories are Apache-2.0 licensed; this derivative snapshot is released under the same license.

This dataset contains synthetic benchmark videos. Users should validate prompts and family-specific evaluators for their intended research setting.