AvrovaDonz commited on
Commit
242b446
·
verified ·
1 Parent(s): 99f09ed

docs(dataset): state the generator-dominated composition and the publication state

Browse files

Records 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": 19319,
17
- "sha256": "66ca8f725a2c6eca830ce8fc50bf9eac294db40d39f6343d914c6f79954ef071"
18
  },
19
  {
20
  "path": "audit/excluded-duplicates.jsonl",
@@ -1263,8 +1263,8 @@
1263
  },
1264
  {
1265
  "path": "tools/build_google_math.py",
1266
- "bytes": 58952,
1267
- "sha256": "b964d358b3aca9eb395d16d3dc00f702ad893d7558520db149dea8867236b614"
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",