Clean schema: parsed candidate JSON, removed redundant columns, added per-row license; Croissant absolute URLs

README.md

@@ -43,13 +43,12 @@ LLMs. Each row is a `(stream, probe, label)` tuple: the model sees a long input
 stream and a short probe, and must answer "yes" or "no" — did the probe occur
 inside the stream?
 
-This release reorganizes the original five-split bundle into three top-level
-splits keyed by modality + source:
+The dataset is organized into three top-level splits keyed by modality + source:
 
 | split | rows | content |
 | --- | --- | --- |
-| `text` | 8,128 | text-modality questions for the synthetic substream benchmark *and* EasyHuman; the `subset` column distinguishes them. |
-| `synthetic_video` | 6,304 | rendered synthetic substream videos *and* EasyHuman 3-belt videos; `subset` distinguishes them. |
+| `text` | 8,128 | text-modality questions for the synthetic substream benchmark *and* EasyHuman; the per-row `split` column distinguishes them as `substream` vs `easyhuman`. |
+| `synthetic_video` | 6,304 | rendered synthetic substream videos *and* EasyHuman 3-belt videos; per-row `split` distinguishes them. |
 | `natural_video` | 1,028 | EPIC-Kitchens-100 derived clips and SoccerNet provenance metadata. |
 
 ## Directory layout
@@ -63,20 +62,13 @@ anonstreammem/substream-recollection/
 │   ├── synthetic-and-easyhuman.txt
 │   └── SoccerNet-NOTE.txt
 ├── text/
-│   ├── questions.parquet         # subset in {substream, easyhuman}
+│   ├── questions.parquet
 │   └── questions.json            # NDJSON copy of the parquet
 ├── synthetic_video/
-│   ├── questions.parquet         # subset in {substream, easyhuman}
+│   ├── questions.parquet
 │   ├── questions.json
 │   └── videos/
-│       ├── L_8_frames/
-│       ├── L_16_frames/
-│       ├── L_32_frames/
-│       ├── L_64_frames/
-│       ├── L_128_frames/
-│       ├── L_256_frames/
-│       ├── L_512_frames/
-│       ├── L_1024_frames/
+│       ├── L_8_frames/ … L_1024_frames/
 │       └── easyhuman_L_256_frames/
 └── natural_video/
     ├── questions.parquet
@@ -118,36 +110,36 @@ root = snapshot_download(
 
 Then `Path(root) / row["video_path"]` resolves to the actual mp4.
 
-## Schema
+## Schema (cleaned 2026-05-03)
 
-All three parquets share most columns. The text and synthetic_video parquets
-also carry `bucket` (original bucket id, e.g. `UNIFORM_EVAL_L0128_ELOW`,
-`len_256`) and `source_split` (the pre-merge split label). `natural_video` adds
-a `source_provenance` JSON column for SoccerNet rows.
+Common columns across all three parquets:
 
-| column | type | notes |
-| --- | --- | --- |
-| `subset` | str | `substream` / `easyhuman` (text + synthetic_video) or bucket id like `L=128` (natural_video) |
-| `bucket` | str | (text + synthetic_video only) original bucket id |
-| `source_split` | str | pre-merge split label |
-| `length_L` | int64 | normalized stream length |
-| `entropy_band` | str | `low` / `medium` / `uniform` / `natural` / `easyhuman` |
-| `modality` | str | `text` or `video` |
-| `question_variant` | str | `sequential` / `spatial` / `binary_natural` / `easyhuman_binary` |
-| `video_index` | int64 | per-bucket video id |
-| `question_index` | int64 | per-video question position |
-| `question_text` | str | prompt the model sees |
-| `candidate` | str (JSON) | probe metadata; null when N/A |
-| `answer` | str | ground-truth label (`yes` / `no`) |
-| `clip_path` | str | relative path to the probe clip |
-| `video_path` | str | relative path to the parent video |
-| `dataset_source` | str | `synthetic` / `easyhuman` / `epic-kitchens-100` / `soccernet` |
-| `input_sequence` | str | comma-joined input symbols (text-mode payload) |
-| `source_class` | str | natural-only event class label |
-| `source_dataset` | str | natural-only origin tag |
-| `video_id` | str | canonical video id |
-| `question_id` | str | canonical question id |
-| `source_provenance` | str (JSON) | natural_video only; SoccerNet provenance |
+- `question_id` (str): canonical question id (note: not globally unique on its own — see "Identifiers" below)
+- `stream_id` (str): id of the parent stream / video this question targets
+- `split` (str): `substream` / `easyhuman` / `natural`
+- `length_L` (int): normalized parent-stream length (8…4096)
+- `entropy_band` (str): `low` / `medium` / `max-entropy` / `easyhuman` / `natural`
+- `question_variant` (str): `sequential` / `spatial` / `easyhuman_binary` / `binary_natural`
+- `question_text` (str): natural-language probe shown to the model
+- `answer` (str): ground-truth `"yes"` or `"no"`
+- `candidate_sequence` (list[str], nullable): probe substream as a list of token strings; null on natural rows
+- `candidate_clip_start` / `candidate_clip_end` (float, nullable): probe-clip time bounds in seconds when applicable
+- `candidate_tag` (str, nullable): EasyHuman category tag (e.g. `x_present`, `mistake_absent`); null on substream/natural
+- `candidate_present` (bool, nullable): ground-truth substream-membership; null on natural
+- `license` (str): per-row license — `CC-BY-4.0`, `CC-BY-NC-4.0`, or `SoccerNet-NDA`
+
+Per-split additional columns:
+
+- `text/questions.parquet`: `input_sequence` (str) — comma-joined input symbols (the text-modality payload)
+- `synthetic_video/questions.parquet`: `video_path` (str), `clip_path` (str) — repo-relative paths
+- `natural_video/questions.parquet`: `video_path` (str), `source_dataset` (str: `epic-kitchens-100` or `soccernet`), `source_class` (str: e.g. `wash_plate`, `red_card`), `source_provenance` (str, JSON: SoccerNet rows only)
+
+### Identifiers
+
+`question_id` is preserved from the source pipeline and is **not** globally
+unique on its own (the same question id appears across multiple buckets/lengths
+when the same probe is reused). The composite key
+`(question_id, length_L, entropy_band, question_variant)` is unique per row.
 
 ## SoccerNet rows
 
@@ -160,12 +152,12 @@ signing the SoccerNet NDA. See `LICENSES/SoccerNet-NOTE.txt`.
 
 ## Licenses
 
-| split / source | license | file |
-| --- | --- | --- |
-| Synthetic streams (`text/subset=substream`, `synthetic_video/subset=substream`) | CC BY 4.0, Anonymous Authors 2026 | `LICENSES/synthetic-and-easyhuman.txt` |
-| EasyHuman (`text/subset=easyhuman`, `synthetic_video/subset=easyhuman`) | CC BY 4.0, Anonymous Authors 2026 | `LICENSES/synthetic-and-easyhuman.txt` |
-| EPIC-Kitchens-100 derived clips (`natural_video`) | CC BY-NC 4.0 (research use only) | `LICENSES/EPIC-Kitchens-100-CC-BY-NC-4.0.txt` |
-| SoccerNet derived rows (`natural_video`, provenance only) | NDA-gated source; not redistributed | `LICENSES/SoccerNet-NOTE.txt` |
+| split / source | license | per-row tag | file |
+| --- | --- | --- | --- |
+| Synthetic streams (`text/split=substream`, `synthetic_video/split=substream`) | CC BY 4.0 | `CC-BY-4.0` | `LICENSES/synthetic-and-easyhuman.txt` |
+| EasyHuman (`text/split=easyhuman`, `synthetic_video/split=easyhuman`) | CC BY 4.0 | `CC-BY-4.0` | `LICENSES/synthetic-and-easyhuman.txt` |
+| EPIC-Kitchens-100 derived clips (`natural_video`, `source_dataset=epic-kitchens-100`) | CC BY-NC 4.0 (research use only) | `CC-BY-NC-4.0` | `LICENSES/EPIC-Kitchens-100-CC-BY-NC-4.0.txt` |
+| SoccerNet derived rows (`natural_video`, `source_dataset=soccernet`; provenance only) | NDA-gated source; not redistributed | `SoccerNet-NDA` | `LICENSES/SoccerNet-NOTE.txt` |
 
 The top-level `license: cc-by-4.0` tag on this card refers to the
 Anonymous-Authors-owned splits (synthetic + EasyHuman) and to this card,
@@ -174,15 +166,3 @@ metadata, and code only. EPIC-Kitchens-100 derivatives remain CC BY-NC 4.0.
 ## Citation
 
 Anonymous, "Substream Recollection," 2026 (anonymized for review).
-
-<!-- SCHEMA_AUTO_BEGIN -->
-
-## Per-row schema (cleaned 2026-05-03)
-
-**`text/questions.parquet`**: `question_id`, `stream_id`, `split`, `length_L`, `entropy_band`, `question_variant`, `question_text`, `answer`, `candidate_sequence`, `candidate_clip_start`, `candidate_clip_end`, `candidate_tag`, `candidate_present`, `input_sequence`, `license`
-
-**`synthetic_video/questions.parquet`**: `question_id`, `stream_id`, `split`, `length_L`, `entropy_band`, `question_variant`, `question_text`, `answer`, `candidate_sequence`, `candidate_clip_start`, `candidate_clip_end`, `candidate_tag`, `candidate_present`, `video_path`, `clip_path`, `license`
-
-**`natural_video/questions.parquet`**: `question_id`, `stream_id`, `split`, `length_L`, `entropy_band`, `question_variant`, `question_text`, `answer`, `candidate_sequence`, `candidate_clip_start`, `candidate_clip_end`, `candidate_tag`, `candidate_present`, `video_path`, `source_dataset`, `source_class`, `source_provenance`, `license`
-
-<!-- SCHEMA_AUTO_END -->

natural_video/questions.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2cf35973ba35065a54d244a49287445a0f29009d462c469bec4450fb58384f44
-size 22823
+oid sha256:51c2b24788be55cb05a9ccf01141941ac17033b4b5ceccd9bcf2842f787aa0ca
+size 24798

synthetic_video/questions.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ee822ae6d5bc89c9721bf5cd3baa075731f7bda039737c3878aebeb9b7ffbae
-size 79243
+oid sha256:73b99a8583857f34f01c54bc350b5fd0079c33ff72f4027ef328d08b085b15cc
+size 111881

text/questions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd873ae313653127cea301f765bb0d90763766dd34e85610b1899dce5c817a62
-size 18594461
+oid sha256:66446f896b96ad40f6a272884c48629c5a862c75fbb79fef06f5bb56cd6cc263
+size 18908001

text/questions.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:912f5bb4f53051518c4e40c5e5d85ffce63d6e2b68aefe61e3a539ebf68f6744
-size 319715
+oid sha256:768e0dcc4542c9eaba744da8882ccea21b90df99b18e8d10b7e4c1e80f673ec8
+size 367170