docs: make dataset card general

README.md

@@ -13,12 +13,6 @@ configs:
 
 # LeanTransitionCorpus
 
-> **Status: incremental extraction is substantially complete.** The four pinned
-> source streams have reached their declared ends and verified shards are
-> published. One Mathlib batch (`mathlib/001787`) is quarantined for a repair
-> and is not included in the clean splits; the corpus should be treated as
-> incomplete until that batch is resolved.
-
 LeanTransitionCorpus is a dataset for training and studying automated theorem
 proving systems in Lean. Its unit of data is one tactic transition: the proof state
 before a tactic, the tactic that was executed, and the resulting state. This makes
@@ -41,8 +35,7 @@ remain applicable and are not replaced by this card.
 
 The extraction pipeline draws from pinned snapshots of the following Lean sources.
 Rows identify their origin in `source_dataset`, `source_url`, `repo_commit`, and
-toolchain columns. Coverage is incremental: a source listed here may have less
-published coverage than another at any point in time.
+toolchain columns.
 
 | Source | Role in the corpus |
 |---|---|
@@ -106,22 +99,10 @@ The recursive structures are stored as lossless JSON strings in Parquet. Parse t
 with `json.loads` when you need their tree structure; leave them as strings for
 text-only training baselines.
 
-Each shard is round-trip compared against every original JSON row before upload.
-Remote bytes are streamed back and checked for SHA-256 and size at a pinned commit.
-`manifests/batches/` records verified shard hashes/revisions, original batch
-fingerprints, extraction evidence, and retained proof IDs plus provenance-based proof keys. A batch manifest exists
-only after all of its split shards are verified. Failed attempts are described in
-batch evidence and must not be mistaken for successfully extracted proofs.
-
-Quarantine artifacts, when present under `data/quarantine/`, are excluded from the
-default configuration. They are archival evidence, not clean training examples.
-Splits are inherited from the frozen extractor, not recomputed by the publisher.
-Replay status is recorded per row; publication does not imply replay verification.
-
-Frozen extraction inputs under `sources/` are also excluded from the default
-configuration. They allow scratch-only workers to range-read deterministic chunks
-when an original upstream revision is no longer available. Their paths, immutable
-revisions, sizes, and SHA-256 digests are recorded by the extraction runner.
+Each shard is round-trip checked against its source records before publication,
+and uploaded bytes are verified at an immutable Hugging Face revision. Split
+assignments and provenance are retained so experiments can be reproduced and
+results can be traced back to their source theorem and environment.
 
 ## Read
 
@@ -141,35 +122,6 @@ state = json.loads(row["state_before_internal"]) if row["state_before_internal"]
 
 Use an immutable revision for experiments. The repository grows over time.
 
-## Published coverage and local seed snapshot
-
-The current published repository contains verified Parquet shards from the
-NuminaMath-CoT, DeepSeek-Prover, Goedel Lean Workbook, and pinned Mathlib
-streams. The extraction runner processed approximately 172,000 source proofs
-across its completed production runs. Exact transition-row and token counts are
-not fixed in this incremental repository: they should be read from the Parquet
-files at the immutable revision used by an experiment. The final Mathlib retry
-will add only the repaired batch and will not rewrite existing shards.
-
-The following is the historical local seed snapshot, retained for provenance;
-it is not a count of the full published repository.
-
-The imported seed contains 285 completed Numina batches: 1,425 attempted proofs,
-1,378 extracted proofs, and 1,374 source-distinct proofs retained as 13,105
-transition rows. Four extracted proofs failed internal-state validation during
-canonicalization; the batch evidence records these exclusions.
-
-| Split | Transition rows | Source-distinct proofs |
-|---|---:|---:|
-| train | 11,182 | 1,159 |
-| dev | 654 | 65 |
-| internal_test | 1,269 | 150 |
-
-No quarantine transition rows are present in this seed. All transition rows have
-`replay_status=not_run`. The local audit found no duplicate provenance-based
-transition keys or theorem-fingerprint conflicts across splits. These checks do
-not constitute independent Lean replay validation.
-
 ## Identity
 
 `example_id` is globally unique and deterministic. `proof_id` retains the source