Datasets:
docs(dataset): state the generator-dominated composition and the publication state
Browse filesRecords the measured record mix in the component README, adds the no-verification caveat next to the composition, points at audit/publish-check.json, and refreshes manifest.json and validation.json for the edited README.
components/google-math/README.md
CHANGED
|
@@ -199,19 +199,37 @@ against their JSONL records, and recomputed group separation from `data/*.jsonl.
|
|
| 199 |
0 spanning two splits, 3,333/35/35. Every measured record, group, character and packed count matched
|
| 200 |
`statistics.json`.
|
| 201 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 202 |
## Limitations
|
| 203 |
|
| 204 |
- No Lean compilation, proof checking, Comparator run or mathematical verification of any kind.
|
|
|
|
|
|
|
|
|
|
| 205 |
- PDF prose proofs are noisy pypdf text; symbol and ligature fidelity is not guaranteed.
|
| 206 |
- The generator subset is a seeded sample of one curriculum, not the paper's full dataset, and its
|
| 207 |
-
questions are school-level rather than research mathematics.
|
|
|
|
|
|
|
| 208 |
- `formal-conjectures` upstream states that some content derives from third-party sources (Wikipedia,
|
| 209 |
MathOverflow and OEIS under CC-BY-SA 4.0). Per-file reference URLs are retained inside the Lean text; no
|
| 210 |
separate per-file license audit was performed.
|
| 211 |
-
- Near-duplicate detection, Phase 1 overlap analysis and editorial review were not performed.
|
|
|
|
| 212 |
- The SentencePiece model normalises runs of spaces, so packed text decodes whitespace-normalised rather
|
| 213 |
than byte-identically to the sources.
|
| 214 |
-
- Only problems that upstream published are present, so several subsets are selection-biased.
|
|
|
|
|
|
|
| 215 |
|
| 216 |
## Statistics
|
| 217 |
|
|
|
|
| 199 |
0 spanning two splits, 3,333/35/35. Every measured record, group, character and packed count matched
|
| 200 |
`statistics.json`.
|
| 201 |
|
| 202 |
+
## Publication
|
| 203 |
+
|
| 204 |
+
The component was first delivered as Hugging Face dataset pull request
|
| 205 |
+
[MMLA-ORG/Cayley-Math discussion #3](https://huggingface.co/datasets/MMLA-ORG/Cayley-Math/discussions/3)
|
| 206 |
+
on ref `refs/pr/3`. Consult that discussion for its current merge state.
|
| 207 |
+
|
| 208 |
+
Delivery needs an account with write access to `MMLA-ORG/Cayley-Math`. An account that can only read the
|
| 209 |
+
repository can open the pull request but cannot complete it, and direct commits to `main` are rejected with
|
| 210 |
+
`403 Forbidden: You have read access but not the required permissions for this operation`.
|
| 211 |
+
|
| 212 |
## Limitations
|
| 213 |
|
| 214 |
- No Lean compilation, proof checking, Comparator run or mathematical verification of any kind.
|
| 215 |
+
- The build ran in a sandbox-free subprocess only for the generator; `cryptids`, `formal-conjectures`,
|
| 216 |
+
`formal-imo`, `formal-putnam-like`, `alphageometry`, `alphaproof-nexus` and the notebooks were never
|
| 217 |
+
built or executed, only read as text.
|
| 218 |
- PDF prose proofs are noisy pypdf text; symbol and ligature fidelity is not guaranteed.
|
| 219 |
- The generator subset is a seeded sample of one curriculum, not the paper's full dataset, and its
|
| 220 |
+
questions are school-level rather than research mathematics. One module, `algebra__linear_1d`, stopped
|
| 221 |
+
at 3,055 of its 8,000 quota when upstream's own `assert entropy == 0` fired; nothing was repaired, and
|
| 222 |
+
the shortfall is recorded in `audit/mathematics-dataset-generation.json`.
|
| 223 |
- `formal-conjectures` upstream states that some content derives from third-party sources (Wikipedia,
|
| 224 |
MathOverflow and OEIS under CC-BY-SA 4.0). Per-file reference URLs are retained inside the Lean text; no
|
| 225 |
separate per-file license audit was performed.
|
| 226 |
+
- Near-duplicate detection, Phase 1 overlap analysis and editorial review were not performed. 1,629
|
| 227 |
+
exact duplicate texts were dropped and are listed in `audit/excluded-duplicates.jsonl`.
|
| 228 |
- The SentencePiece model normalises runs of spaces, so packed text decodes whitespace-normalised rather
|
| 229 |
than byte-identically to the sources.
|
| 230 |
+
- Only problems that upstream published are present, so several subsets are selection-biased. `cryptids`
|
| 231 |
+
contributes 31.6M of the 78.4M characters because a handful of Busy Beaver traces are enormous, so the
|
| 232 |
+
corpus is dominated by machine-generated proof script rather than mathematics prose.
|
| 233 |
|
| 234 |
## Statistics
|
| 235 |
|
components/google-math/manifest.json
CHANGED
|
@@ -13,8 +13,8 @@
|
|
| 13 |
},
|
| 14 |
{
|
| 15 |
"path": "README.md",
|
| 16 |
-
"bytes":
|
| 17 |
-
"sha256": "
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"path": "audit/excluded-duplicates.jsonl",
|
|
@@ -1263,8 +1263,8 @@
|
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"path": "tools/build_google_math.py",
|
| 1266 |
-
"bytes":
|
| 1267 |
-
"sha256": "
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
"path": "tools/generate_mathematics_dataset_sample.py",
|
|
|
|
| 13 |
},
|
| 14 |
{
|
| 15 |
"path": "README.md",
|
| 16 |
+
"bytes": 20658,
|
| 17 |
+
"sha256": "78949809ffc4dade580bec22617b9a20146c29869b7b55ee7aa00eac0f0a7edc"
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"path": "audit/excluded-duplicates.jsonl",
|
|
|
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"path": "tools/build_google_math.py",
|
| 1266 |
+
"bytes": 60423,
|
| 1267 |
+
"sha256": "277b2505e62ffc8d552d398c55fbc7bf9d9e00d5fb2099109b7bfae5b5bbf942"
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
"path": "tools/generate_mathematics_dataset_sample.py",
|