Initial dataset upload (anonymized for review)

.gitattributes

@@ -1,6 +1,5 @@
 *.7z filter=lfs diff=lfs merge=lfs -text
 *.arrow filter=lfs diff=lfs merge=lfs -text
-*.avro filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
 *.bz2 filter=lfs diff=lfs merge=lfs -text
 *.ckpt filter=lfs diff=lfs merge=lfs -text
@@ -58,3 +57,14 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+nodes.csv filter=lfs diff=lfs merge=lfs -text
+edges.csv filter=lfs diff=lfs merge=lfs -text
+mathlib_nodes.csv filter=lfs diff=lfs merge=lfs -text
+mathlib_edges.csv filter=lfs diff=lfs merge=lfs -text
+tactic_usage.ndjson filter=lfs diff=lfs merge=lfs -text
+mechanisms.ndjson filter=lfs diff=lfs merge=lfs -text
+v2/mathlib_nodes_enriched.csv filter=lfs diff=lfs merge=lfs -text
+v2/mathlib_namespace_edges_k2.csv filter=lfs diff=lfs merge=lfs -text
+v2/declaration/nodes.csv filter=lfs diff=lfs merge=lfs -text
+v2/namespace/edges.csv filter=lfs diff=lfs merge=lfs -text
+v2/declaration/metrics.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,232 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+task_categories:
+  - graph-ml
+language:
+  - en
+tags:
+  - mathematics
+  - lean4
+  - mathlib
+  - dependency-graph
+  - formal-verification
+  - network-analysis
+size_categories:
+  - 100K<n<1M
+---
+
+# MathlibGraph: The Multinetwork of Mathlib
+
+Dependency graph of [Mathlib](https://github.com/leanprover-community/mathlib4) (commit [`534cf0b`](https://github.com/leanprover-community/mathlib4/commit/534cf0b), 2 Feb 2026), the largest formal mathematics library for Lean 4 (`v4.28.0-rc1`).
+
+Three dependency layers (declarations, modules, namespaces), each with nodes, edges, and precomputed network metrics.
+
+## Quick Stats
+
+| | Declarations | Modules | Namespaces (k=2) |
+|---|---|---|---|
+| Nodes | 308,129 | 7,564 | 10,097 |
+| Edges | 8,436,366 | 20,881 | 332,081 (weighted) |
+| DAG depth | 83 | 154 | 7 (after SCC condensation) |
+| Louvain modularity | 0.478 | 0.610 | 0.270 |
+| Synthesized edges | 74.2% | — | — |
+| In cycles | 5,732 nodes | 0 (DAG) | 6,055 nodes (38 SCCs) |
+
+## Files
+
+### Raw Data (v1)
+
+| File | Rows | Description |
+|------|------|-------------|
+| `mathlib_edges.csv` | 8,436,366 | Declaration dependency edges |
+| `mathlib_nodes.csv` | 317,655 | Declarations before deduplication |
+| `nodes.csv` / `edges.csv` | 633K / 10.9M | Full environment (Lean + Std + Mathlib) |
+| `mechanisms.ndjson` | — | Lean language mechanism extractions |
+| `tactic_usage.ndjson` | — | Per-declaration tactic usage profiles |
+
+### Three-Layer Graphs (v2)
+
+Each layer has `nodes.csv`, `edges.csv`, and `metrics.csv`:
+
+| Folder | Nodes | Edges | Metrics | Description |
+|--------|-------|-------|---------|-------------|
+| `v2/declaration/` | 308,129 | (use root `mathlib_edges.csv`) | 308,129 x 11 | Theorems, definitions, and other named constants |
+| `v2/module/` | 7,564 | 20,881 | 7,564 x 10 | Source files linked by `import` statements |
+| `v2/namespace/` | 10,097 | 332,081 | 10,097 x 11 | Depth-2 dotted-name prefixes, weighted edges |
+
+`v2/summary.json` contains headline statistics for all three levels.
+
+## Quick Start
+
+```python
+from datasets import load_dataset
+
+# Declaration level
+decl = load_dataset("anon-d3a4/MathlibGraph",
+                    data_files="v2/declaration/metrics.csv", split="train").to_pandas()
+edges = load_dataset("anon-d3a4/MathlibGraph",
+                     data_files="mathlib_edges.csv", split="train").to_pandas()
+
+# Module level
+mod_nodes = load_dataset("anon-d3a4/MathlibGraph",
+                         data_files="v2/module/nodes.csv", split="train").to_pandas()
+mod_edges = load_dataset("anon-d3a4/MathlibGraph",
+                         data_files="v2/module/edges.csv", split="train").to_pandas()
+
+# Namespace level
+ns_nodes = load_dataset("anon-d3a4/MathlibGraph",
+                        data_files="v2/namespace/nodes.csv", split="train").to_pandas()
+ns_edges = load_dataset("anon-d3a4/MathlibGraph",
+                        data_files="v2/namespace/edges.csv", split="train").to_pandas()
+```
+
+## Schema
+
+### v2/declaration/nodes.csv
+
+Deduplicated from `mathlib_nodes.csv` (317,655 to 308,129 rows; 9,526 `@[to_additive]` mirrors removed).
+
+| Column | Type | Description |
+|--------|------|-------------|
+| name | str | Fully qualified declaration name |
+| kind | str | theorem, definition, abbrev, inductive, constructor, opaque, axiom, or quotient |
+| module | str | Parent namespace (null for 30,944 Lean core declarations) |
+
+### v2/declaration/metrics.csv
+
+All columns from `nodes.csv` plus:
+
+| Column | Type | Description |
+|--------|------|-------------|
+| namespace_depth2 | str | First 2 dot-separated components (e.g., `Mathlib.Algebra`) |
+| namespace_depth3 | str | First 3 dot-separated components |
+| in_degree | int | Declarations that depend on this one |
+| out_degree | int | Declarations this one depends on |
+| pagerank | float | PageRank (alpha=0.85) |
+| betweenness | float | Betweenness centrality (k=500, seed=42) |
+| community_id | int | Louvain community |
+| dag_layer | int | Topological depth; -1 for 5,732 nodes in cycles |
+| file_module | str | Source file module (from Lean environment, 278K coverage) |
+| is_tactic_proof | bool | Whether the proof uses tactics (78,315 declarations) |
+| tactic_count | int | Number of tactics used in the proof |
+| top_tactic | str | Most frequently used tactic in this proof |
+| is_instance | bool | Whether this is a typeclass instance (26,415 declarations) |
+| instance_class | str | Which typeclass this instance implements |
+| is_coercion | bool | Whether this is a coercion (241 declarations) |
+| to_additive_pair | str | Name of the corresponding additive/multiplicative variant |
+| def_height | float | Definitional height in the kernel (42,935 declarations) |
+
+### v2/module/nodes.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| module | str | Dotted module name (e.g., `Mathlib.Algebra.Group.Defs`) |
+| decl_count | int | Declarations defined in this source file |
+
+### v2/module/edges.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| source | str | Importing module |
+| target | str | Imported module |
+| is_exported | bool | Whether this is a public import (20,699 of 20,881) |
+
+### v2/module/metrics.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| module | str | Module name |
+| decl_count | int | Declarations in this module (full decl_module mapping, 499K coverage) |
+| in_degree | int | Modules that import this one |
+| out_degree | int | Modules this one imports |
+| pagerank | float | PageRank on module import graph |
+| betweenness | float | Betweenness centrality (exact) |
+| dag_layer | int | Topological depth in module DAG |
+| community_id | int | Louvain community |
+| cohesion | float | Fraction of declaration edges staying within module |
+| import_utilization_median | float | Median fraction of imported declarations used |
+
+### v2/namespace/nodes.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| namespace | str | Depth-2 namespace (e.g., `Mathlib.Algebra`) |
+| decl_count | int | Declarations in this namespace |
+| in_cycle | bool | True if this namespace is in a strongly connected component |
+| scc_id | int | SCC identifier (-1 if not in a cycle) |
+
+### v2/namespace/edges.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| source | str | Source namespace |
+| target | str | Target namespace |
+| weight | int | Number of declaration-level edges between these namespaces |
+
+### v2/namespace/metrics.csv
+
+| Column | Type | Description |
+|--------|------|-------------|
+| namespace | str | Depth-2 namespace |
+| decl_count | int | Declarations in this namespace |
+| in_degree | int | Unweighted in-degree |
+| out_degree | int | Unweighted out-degree |
+| edge_weight_sum | int | Total declaration edges involving this namespace |
+| pagerank | float | Weighted PageRank (alpha=0.85) |
+| betweenness | float | Weighted betweenness (k=300, seed=42) |
+| community_id | int | Louvain community (weighted undirected) |
+| cross_ns_ratio | float | Fraction of edges crossing namespace boundaries |
+| in_cycle | bool | In a strongly connected component (6,055 of 10,097) |
+| scc_id | int | SCC identifier (-1 if acyclic) |
+
+## Methodology
+
+- **Extraction**: lean4export (nodes) + lean-training-data (edges) + importGraph (module graph) + jixia (metadata)
+- **Deduplication**: by name, 317,655 to 308,129 rows (9,526 `@[to_additive]` mirrors)
+- **Self-loops**: 4,755 constructor self-references filtered
+- **PageRank**: alpha=0.85, max_iter=100, tol=1e-6
+- **Betweenness**: declaration k=500, namespace k=300, module exact; seed=42
+- **Communities**: Louvain, resolution=1.0, random_state=42, undirected projection
+- **DAG layers**: Kahn's algorithm; cycle nodes get layer=-1; namespace graph condensed via SCC
+- **Namespace cycles**: 6,055 of 10,097 namespaces in 38 SCCs (largest: 5,899 nodes)
+
+## Hold-Out Experiments
+
+We validate the premise retrieval results with two levels of hold-out experiments to assess information leakage. In the original experiment, all network features (degree, PageRank, betweenness, community, DAG layer) are precomputed on the full graph. Hold-out experiments recompute features on a reduced graph to test whether full-graph computation inflates AUC.
+
+### Edge-level hold-out
+- Randomly remove 20% of edges, recompute all features on remaining 80% graph
+- Tests whether full-graph feature computation inflates AUC
+
+### Declaration-level hold-out
+- Split theorems 80/20, remove ALL edges (incoming and outgoing) of test theorems from training graph
+- Test theorems have zero degree in training graph
+- Simulates the real scenario: predicting premises for a brand new theorem
+
+### Results (Split 1, seed=42)
+
+| Method | AUC (original) | AUC (edge hold-out) | AUC (decl hold-out) |
+|--------|---------------|---------------------|---------------------|
+| Random | 0.499 | 0.497 | 0.500 |
+| Same module | 0.563 | 0.562 | 0.563 |
+| Same namespace | 0.590 | 0.590 | 0.592 |
+| Same community | 0.768 | 0.755 | 0.500 |
+| Network features | 0.991 | 0.988 | 0.978 |
+| All features | 0.994 | 0.992 | 0.984 |
+
+The community feature drops to random (0.500) in declaration-level hold-out because test declarations become isolated nodes with unique community IDs. Despite this, the combined network feature model retains AUC=0.978, confirming that degree, PageRank, and DAG position carry genuine predictive signal independent of information leakage.
+
+Full 5-split results with mean and std are in `experiments/holdout_decl_level.json` (updated incrementally).
+
+### Experiment Files
+
+| File | Description |
+|------|-------------|
+| `experiments/premise_retrieval_results.json` | Original experiment: 6 methods, 4 metrics, 95% CI |
+| `experiments/premise_retrieval_hard_negatives.json` | Hard negatives (same-community) variant |
+| `experiments/holdout_edge_level.json` | Edge-level hold-out (1 split) |
+| `experiments/holdout_decl_level.json` | Declaration-level hold-out (5 splits, incremental) |
+
+## License
+
+Apache 2.0

edges.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78b20d6311388159bdab03ddfb68d5ef5687ced629ee04a99d9880bcd043a08f
+size 753711915

experiments/holdout_decl_level.json

@@ -0,0 +1,113 @@
+{
+  "completed_splits": [
+    {
+      "seed": 42,
+      "results_uniform": {
+        "Random": {
+          "AUC": 0.5003,
+          "R@10": 0.1333,
+          "R@50": 0.676,
+          "MRR": 0.524,
+          "n_problems": 5000
+        },
+        "Same module": {
+          "AUC": 0.563,
+          "R@10": 0.1789,
+          "R@50": 0.642,
+          "MRR": 0.998,
+          "n_problems": 5000
+        },
+        "Same namespace": {
+          "AUC": 0.5922,
+          "R@10": 0.2233,
+          "R@50": 0.664,
+          "MRR": 0.966,
+          "n_problems": 5000
+        },
+        "Same community": {
+          "AUC": 0.5,
+          "R@10": 0.0683,
+          "R@50": 0.618,
+          "MRR": 1.0,
+          "n_problems": 5000
+        },
+        "Network features": {
+          "AUC": 0.9778,
+          "R@10": 0.5124,
+          "R@50": 0.937,
+          "MRR": 0.997,
+          "n_problems": 5000
+        },
+        "All features": {
+          "AUC": 0.9844,
+          "R@10": 0.5212,
+          "R@50": 0.94,
+          "MRR": 0.997,
+          "n_problems": 5000
+        }
+      },
+      "results_hard": {
+        "Random": {
+          "AUC": 0.5003,
+          "R@10": 0.1333,
+          "R@50": 0.676,
+          "MRR": 0.524,
+          "n_problems": 5000
+        },
+        "Same module": {
+          "AUC": 0.563,
+          "R@10": 0.1789,
+          "R@50": 0.642,
+          "MRR": 0.997,
+          "n_problems": 5000
+        },
+        "Same namespace": {
+          "AUC": 0.5922,
+          "R@10": 0.2234,
+          "R@50": 0.664,
+          "MRR": 0.964,
+          "n_problems": 5000
+        },
+        "Same community": {
+          "AUC": 0.5,
+          "R@10": 0.0683,
+          "R@50": 0.618,
+          "MRR": 1.0,
+          "n_problems": 5000
+        },
+        "Network features": {
+          "AUC": 0.9776,
+          "R@10": 0.5097,
+          "R@50": 0.937,
+          "MRR": 0.997,
+          "n_problems": 5000
+        },
+        "All features": {
+          "AUC": 0.9842,
+          "R@10": 0.5191,
+          "R@50": 0.94,
+          "MRR": 0.997,
+          "n_problems": 5000
+        }
+      },
+      "diagnostics": {
+        "in_degree_pearson_r": 0.999851,
+        "out_degree_pearson_r": 0.999315,
+        "pagerank_pearson_r": 0.950683,
+        "betweenness_pearson_r": 0.712134,
+        "community_nmi": 0.7821,
+        "dag_layer_mean_shift": 2.9912,
+        "test_decl_mean_in_degree": 0.0,
+        "test_decl_mean_out_degree": 0.0,
+        "test_decl_mean_pagerank": 7.667e-07,
+        "n_wcc": 48812,
+        "largest_wcc_size": 259311,
+        "g_train_edges": 6727267,
+        "edges_removed_pct": 20.21
+      },
+      "n_train_decls": 194929,
+      "n_test_decls": 48733,
+      "elapsed_s": 2396
+    }
+  ]
+}

experiments/holdout_edge_level.json

@@ -0,0 +1,63 @@
+{
+  "completed_splits": [
+    {
+      "seed": 42,
+      "results": {
+        "Random": {
+          "AUC": 0.497,
+          "R@10": 0.179,
+          "R@50": 0.897,
+          "MRR": 0.25,
+          "n_problems": 5000
+        },
+        "Same module": {
+          "AUC": 0.562,
+          "R@10": 0.398,
+          "R@50": 0.973,
+          "MRR": 0.994,
+          "n_problems": 5000
+        },
+        "Same namespace": {
+          "AUC": 0.59,
+          "R@10": 0.431,
+          "R@50": 0.974,
+          "MRR": 0.93,
+          "n_problems": 5000
+        },
+        "Same community": {
+          "AUC": 0.755,
+          "R@10": 0.441,
+          "R@50": 0.98,
+          "MRR": 0.786,
+          "n_problems": 5000
+        },
+        "Network features": {
+          "AUC": 0.988,
+          "R@10": 0.919,
+          "R@50": 0.999,
+          "MRR": 0.976,
+          "n_problems": 5000
+        },
+        "All features": {
+          "AUC": 0.992,
+          "R@10": 0.926,
+          "R@50": 0.999,
+          "MRR": 0.984,
+          "n_problems": 5000
+        }
+      },
+      "diagnostics": {
+        "in_degree_pearson_r": 0.999991,
+        "out_degree_pearson_r": 0.996016,
+        "pagerank_pearson_r": 0.911671,
+        "betweenness_pearson_r": 0.658761,
+        "community_nmi": 0.6737,
+        "dag_layer_mean_shift": 3.1637,
+        "dag_layer_max_shift": 80,
+        "n_wcc": 363,
+        "largest_wcc_size": 307763
+      },
+      "elapsed_s": 2937
+    }
+  ]
+}

experiments/premise_retrieval_hard_negatives.json

@@ -0,0 +1,38 @@
+{
+  "Random": {
+    "AUC": 0.5015032262104209,
+    "R@10": 0.1360585603981675,
+    "R@50": 0.6769906958015243,
+    "MRR": 0.5216367807667213
+  },
+  "Same module": {
+    "AUC": 0.562350943150883,
+    "R@10": 0.17716134828405006,
+    "R@50": 0.6429486391917959,
+    "MRR": 0.9893142857142857
+  },
+  "Same namespace": {
+    "AUC": 0.5803446516951352,
+    "R@10": 0.2057242294389468,
+    "R@50": 0.6591106708054222,
+    "MRR": 0.9167673885966564
+  },
+  "Same community": {
+    "AUC": 0.3601798080569178,
+    "R@10": 0.05207239139664216,
+    "R@50": 0.5079945224060302,
+    "MRR": 0.793327165399728
+  },
+  "Network features": {
+    "AUC": 0.9838247131236855,
+    "R@10": 0.5103782600858937,
+    "R@50": 0.9414442050593969,
+    "MRR": 0.9986583333333334
+  },
+  "All features": {
+    "AUC": 0.9894400400671409,
+    "R@10": 0.5203356738855442,
+    "R@50": 0.9426993462065937,
+    "MRR": 0.9986333333333334
+  }
+}

experiments/premise_retrieval_results.json

@@ -0,0 +1,86 @@
+{
+  "Random": {
+    "AUC": 0.4988464624381927,
+    "R@10": 0.13489873938604055,
+    "R@50": 0.675421720785031,
+    "MRR": 0.5177648872389418,
+    "AUC_ci_lo": 0.4976579056909061,
+    "AUC_ci_hi": 0.5025447127337402,
+    "R@10_ci_lo": 0.13314589782962657,
+    "R@10_ci_hi": 0.13807478534717094,
+    "R@50_ci_lo": 0.6708032287760948,
+    "R@50_ci_hi": 0.6822757937354658,
+    "MRR_ci_lo": 0.5097106183497658,
+    "MRR_ci_hi": 0.5304948072004493
+  },
+  "Same module": {
+    "AUC": 0.563252943150883,
+    "R@10": 0.17701251024723705,
+    "R@50": 0.6434465696480249,
+    "MRR": 0.9972,
+    "AUC_ci_lo": 0.5611102003189763,
+    "AUC_ci_hi": 0.5651486253506915,
+    "R@10_ci_lo": 0.17269307941264872,
+    "R@10_ci_hi": 0.18132900160478696,
+    "R@50_ci_lo": 0.6354730399715295,
+    "R@50_ci_hi": 0.6551981898656606,
+    "MRR_ci_lo": 0.9961949999999999,
+    "MRR_ci_hi": 0.9981
+  },
+  "Same namespace": {
+    "AUC": 0.5902006516951351,
+    "R@10": 0.2200081147938892,
+    "R@50": 0.6626223704593136,
+    "MRR": 0.9695453174603176,
+    "AUC_ci_lo": 0.5880529846813214,
+    "AUC_ci_hi": 0.5924840785351329,
+    "R@10_ci_lo": 0.21586994086758424,
+    "R@10_ci_hi": 0.22475987434791297,
+    "R@50_ci_lo": 0.6548427218687078,
+    "R@50_ci_hi": 0.6734164434880273,
+    "MRR_ci_lo": 0.9664740277777778,
+    "MRR_ci_hi": 0.9724816488095238
+  },
+  "Same community": {
+    "AUC": 0.7675738080569178,
+    "R@10": 0.2066798209803219,
+    "R@50": 0.8225669566396177,
+    "MRR": 0.889924414802174,
+    "AUC_ci_lo": 0.7636466473774499,
+    "AUC_ci_hi": 0.7701280102171949,
+    "R@10_ci_lo": 0.20206740759955272,
+    "R@10_ci_hi": 0.2099666022599327,
+    "R@50_ci_lo": 0.8168676201838171,
+    "R@50_ci_hi": 0.8301017721564548,
+    "MRR_ci_lo": 0.8825203088999172,
+    "MRR_ci_hi": 0.8968358562953322
+  },
+  "Network features": {
+    "AUC": 0.9910488074402863,
+    "R@10": 0.5200779274690902,
+    "R@50": 0.9440300907986493,
+    "MRR": 0.9978666666666666,
+    "AUC_ci_lo": 0.9906065336840136,
+    "AUC_ci_hi": 0.9915709117833958,
+    "R@10_ci_lo": 0.513202566369347,
+    "R@10_ci_hi": 0.5287072912456959,
+    "R@50_ci_lo": 0.9401626991702056,
+    "R@50_ci_hi": 0.9492015962873687,
+    "MRR_ci_lo": 0.9969158333333333,
+    "MRR_ci_hi": 0.9986
+  },
+  "All features": {
+    "AUC": 0.9935149509905005,
+    "R@10": 0.5246824912435099,
+    "R@50": 0.944575070939189,
+    "MRR": 0.9984,
+    "AUC_ci_lo": 0.9931381533391922,
+    "AUC_ci_hi": 0.9938923361368286,
+    "R@10_ci_lo": 0.51749162029253,
+    "R@10_ci_hi": 0.5334405006567039,
+    "R@50_ci_lo": 0.9407193927667646,
+    "R@50_ci_hi": 0.9497068095655786,
+    "MRR_ci_lo": 0.9975475,
+    "MRR_ci_hi": 0.999
+  }
+}

mathlib_edges.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e1426b2304a97924cca1e48fa22d76620e1cb7bbb496a1bd9dd742cc6906922
+size 584463744

mathlib_nodes.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b222647e7f5871ae20603c2ef87ec0da58143f097e03babdd584c654061725cb
+size 17813672

mechanisms.ndjson

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8d0905876f58af256e2f56bc6ac3c8f90f6a6e5d4c0a4b7f921e3e113e123f9
+size 86007387

nodes.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf2578bb394a14efb77bb5ce31e3313e5f24498a86139aa35be1f2c8c5a685b2
+size 51121872

summary.md

@@ -0,0 +1,46 @@
+# MathFactor Data Summary
+
+**Validation**: WARNING - Some edges reference missing nodes
+  - Missing sources: 15914 (e.g., ['Set.preimage_add_preimage_subset', 'Left.add_neg_of_neg_of_nonpos', 'Hindman.FS.head'])
+  - Missing targets: 12885 (e.g., ['Set.preimage_add_preimage_subset', 'GroupSeminorm.instSupSet._proof_3._to_additive_1', 'dist_sub_eq_dist_add_right'])
+
+## Nodes
+
+**Total**: 633,364
+
+| Kind | Count |
+|------|-------|
+| theorem | 441,230 |
+| definition | 82,962 |
+| abbrev | 79,627 |
+| constructor | 9,070 |
+| opaque | 7,356 |
+| recursor | 6,614 |
+| inductive | 6,494 |
+| axiom | 7 |
+| quotient | 4 |
+
+## Edges
+
+**Total**: 10,889,011
+
+| Type | Count |
+|------|-------|
+| Explicit (*) | 3,248,395 |
+| Simplifier (s) | 0 |
+| Direct | 7,640,616 |
+
+## Top Modules (by node count)
+
+| Module | Count |
+|--------|-------|
+|  | 31,302 |
+| List | 5,233 |
+| Set | 5,065 |
+| Finset | 4,954 |
+| Nat | 4,566 |
+| MeasureTheory | 3,496 |
+| CategoryTheory | 3,471 |
+| Polynomial | 3,280 |
+| Int | 2,929 |
+| CategoryTheory.Limits | 2,757 |

tactic_usage.ndjson

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3927807af5920b626b687565f415233219df017b5b1555a074a073809bff4f84
+size 45050618

v2/declaration/metrics.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c88b5044db57d48ba150fbe0a3550d47684739ca9b97dbb141564c5ce7a5658
+size 57521366

v2/declaration/nodes.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0d2ad9ce65b83b1b650455eebea408b9820e9eef2ac8524d2b70a496044247e
+size 37905777

v2/namespace/edges.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83bfba3f35e645002b03877bd82cf0013b9ee79d858c9a7b6fbe6281981fbc92
+size 10799848

v2/summary.json

@@ -0,0 +1,68 @@
+{
+  "snapshot": {
+    "mathlib_commit": "534cf0b",
+    "snapshot_date": "2026-02-02",
+    "lean_version": "leanprover/lean4:v4.28.0-rc1"
+  },
+  "declaration_graph": {
+    "node_count": 308129,
+    "edge_count_raw": 8436366,
+    "edge_count_no_self_loops": 8431611,
+    "self_loops": 4755,
+    "explicit_edge_count": 2178534,
+    "explicit_edge_ratio": 0.2582,
+    "synthesized_edge_ratio": 0.7418,
+    "dag_depth": 83,
+    "dag_layers": 84,
+    "nodes_in_cycles": 5732,
+    "louvain_modularity": 0.4778,
+    "louvain_num_communities": 89,
+    "nmi_namespace_vs_community": 0.3458,
+    "power_law_alpha_in": 1.781,
+    "power_law_alpha_in_source": "paper appendix-decl.tex table, xmin=20",
+    "power_law_alpha_out": 2.936,
+    "power_law_alpha_out_source": "paper appendix-decl.tex table"
+  },
+  "module_graph": {
+    "node_count": 7564,
+    "edge_count": 20881,
+    "dag_depth": 154,
+    "dag_layers": 155,
+    "louvain_modularity": 0.6101,
+    "louvain_num_communities": 15,
+    "transitive_redundancy_ratio": 0.175,
+    "transitive_redundancy_ratio_source": "paper contribution.tex",
+    "median_import_utilization": 0.016,
+    "median_import_utilization_source": "paper contribution.tex (per-edge median over all import edges)",
+    "critical_path_parallelism_ratio": 22.4,
+    "critical_path_parallelism_ratio_source": "paper appendix-module.tex"
+  },
+  "namespace_graph_k2": {
+    "node_count": 10097,
+    "edge_count": 332081,
+    "dag_depth": 7,
+    "dag_layers": 8,
+    "dag_note": "namespace graph has cycles; depth computed on condensed SCC graph",
+    "louvain_modularity": 0.2703,
+    "louvain_num_communities": 13,
+    "nmi_namespace_vs_community": 0.2456,
+    "cross_namespace_ratio": 0.8581,
+    "cross_namespace_ratio_note": "fraction of declaration edges crossing depth-2 namespace boundaries"
+  },
+  "edge_decompositions": {
+    "proof_only_ratio": 0.439,
+    "statement_only_ratio": 0.081,
+    "mixed_ratio": 0.48,
+    "source": "paper contribution.tex (requires Lean extraction not in current dataset)"
+  },
+  "kind_distribution": {
+    "theorem": 243725,
+    "definition": 48664,
+    "abbrev": 6667,
+    "constructor": 4755,
+    "inductive": 3813,
+    "opaque": 499,
+    "quotient": 3,
+    "axiom": 3
+  }
+}