Datasets:

Modalities:
Text
Formats:
json
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
hanwenzhu commited on
Commit
22d93fb
·
verified ·
1 Parent(s): 0e315c8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +165 -3
README.md CHANGED
@@ -1,3 +1,165 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+
4
+ configs:
5
+ - config_name: carleson
6
+ data_files:
7
+ - split: valid
8
+ path: "minictx-valid/carleson.jsonl"
9
+ - split: test
10
+ path: "minictx-test/carleson.jsonl"
11
+ - config_name: ConNF
12
+ data_files:
13
+ - split: valid
14
+ path: "minictx-valid/ConNF.jsonl"
15
+ - split: test
16
+ path: "minictx-test/ConNF.jsonl"
17
+ - config_name: FLT
18
+ data_files:
19
+ - split: valid
20
+ path: "minictx-valid/FLT.jsonl"
21
+ - split: test
22
+ path: "minictx-test/FLT.jsonl"
23
+ - config_name: foundation
24
+ data_files:
25
+ - split: valid
26
+ path: "minictx-valid/foundation.jsonl"
27
+ - split: test
28
+ path: "minictx-test/foundation.jsonl"
29
+ - config_name: HepLean
30
+ data_files:
31
+ - split: valid
32
+ path: "minictx-valid/HepLean.jsonl"
33
+ - split: test
34
+ path: "minictx-test/HepLean.jsonl"
35
+ - config_name: mathlib
36
+ data_files:
37
+ - split: valid
38
+ path: "minictx-valid/mathlib.jsonl"
39
+ - split: test
40
+ path: "minictx-test/mathlib.jsonl"
41
+ - config_name: Seymour
42
+ data_files:
43
+ - split: valid
44
+ path: "minictx-valid/Seymour.jsonl"
45
+ - split: test
46
+ path: "minictx-test/Seymour.jsonl"
47
+ ---
48
+
49
+ miniCTX-v2 is a context-rich dataset for evaluation of neural theorem proving under a more realistic scenario. [Project page](https://cmu-l3.github.io/minictx/)
50
+
51
+ ## Example Entry
52
+ An entry in the miniCTX dataset consists of the theorem statement, preceding file contents, and metadata information. For example, given the following theorem `s_eq_pow_two` in context:
53
+
54
+ ```lean
55
+ import Mathlib.Data.Real.Basic
56
+
57
+ /-!
58
+ # Square function
59
+ We define the squaring function `s : ℝ → ℝ` to be `s x := x * x`.
60
+ -/
61
+
62
+ def s (x : ℝ) : ℝ := x * x
63
+
64
+ lemma s_eq_pow_two {x : ℝ} : s x = x ^ 2 := by
65
+ rw [s, pow_two]
66
+ ```
67
+
68
+ The problem is formatted in JSON as follows:
69
+
70
+ ```json
71
+ {
72
+ # Preceding file content
73
+ "srcContext": "import Mathlib.Data.Real.Basic\n\n/-!\n# Square function\nWe define the squaring function `s : ℝ → ℝ` to be `s x := x * x`.\n-/\n\ndef s (x : ℝ) : ℝ := x * x\n\n",
74
+
75
+ # Theorem statement
76
+ "theoremStatement": "lemma s_eq_pow_two {x : ℝ} : s x = x ^ 2",
77
+
78
+ # Fully qualified theorem name
79
+ "theoremName": "s_eq_pow_two",
80
+
81
+ # Temporal metadata
82
+ "fileCreated": {"commit":"(git commit)", "date":"(date the commit is updated)"},
83
+ "theoremCreated": {"commit":"(git commit)", "date":"(date the commit is updated)"},
84
+
85
+ # Source metadata
86
+ "file": "MyProject/Square.lean",
87
+ "module": "MyProject.Square",
88
+ "positionMetadata": {
89
+ # Line number the theorem is on
90
+ "lineInFile": 10,
91
+ # Number of tokens before the theorem
92
+ "tokenPositionInFile": 152,
93
+ # Number of premises (definitions, theorems) before the theorem
94
+ "theoremPositionInFile": 1
95
+ },
96
+
97
+ # Dependency metadata
98
+ "dependencyMetadata": {
99
+ # Number of definitions or lemmas defined in this file that the theorem uses
100
+ "inFilePremises": true,
101
+ "numInFilePremises": 1,
102
+ # Number of definitions or lemmas defined in this repository that the theorem uses (including in-file ones)
103
+ "repositoryPremises": true
104
+ "numRepositoryPremises": 1,
105
+ # Number of total premises (in file, repository, or otherwise)
106
+ "numPremises": 2,
107
+ },
108
+
109
+ # Proof metadata
110
+ "proofMetadata": {
111
+ "hasProof": true,
112
+ "proof": "by\n rw [s, pow_two]",
113
+ "proofType": "tactic",
114
+ "proofLengthLines": 2,
115
+ "proofLengthTokens": 20
116
+ }
117
+ }
118
+ ```
119
+
120
+ ### Description of Each Entry
121
+
122
+ - **srcContext**: The context of the source file preceding the theorem, including imports and relevant definitions. This provides necessary background to understand the theorem.
123
+ - **theoremStatement**: The statement of the theorem being proved.
124
+ - **theoremName**: The name of the theorem.
125
+ - **fileCreated**: The git commit hash indicating when the file was created.
126
+ - **theoremCreated**: The git commit hash indicating when the theorem was added.
127
+ - **file**: The name of the file containing the theorem.
128
+ - **positionMetadata**:
129
+ - **lineInFile**: The line number in the file where the theorem is located.
130
+ - **tokenPositionInFile**: The number of tokens in the file before the theorem starts.
131
+ - **theoremPositionInFile**: The number of premises (definitions, theorems) before the theorem in the file.
132
+ - **dependencyMetadata**:
133
+ - **inFilePremises**: Indicates whether the theorem uses definitions or lemmas defined in the same file.
134
+ - **numInFilePremises**: The number of definitions or lemmas from the same file that the theorem relies on.
135
+ - **repositoryPremises**: Indicates whether the theorem uses definitions or lemmas defined in another file within the same repository.
136
+ - **numRepositoryPremises**: The total number of definitions or lemmas from the repository (including the current file) that the theorem depends on.
137
+ - **numPremises**: The total number of premises the theorem depends on, regardless of whether they are from the same file, the repository, or external sources.
138
+ - **proofMetadata**:
139
+ - **hasProof**: Indicates whether the theorem has a proof.
140
+ - **proof**: The proof of the theorem.
141
+ - **proofType**: The type of proof, term proof or tactic proof.
142
+ - **proofLengthLines**: The length of the proof in lines.
143
+ - **proofLengthTokens**: The length of the proof in tokens.
144
+
145
+ In addition to individual entries, we also provide the link and git commit version of each split for evaluation:
146
+ - Carleson: https://github.com/fpvandoorn/carleson, commit a5d265f109105809de4aaff16776b7c16b1c0bd5
147
+ - ConNF: https://github.com/leanprover-community/con-nf, commit 51c38ad244870b8b1f40b8272b281678397dfa4f
148
+ - FLT: https://github.com/ImperialCollegeLondon/FLT, commit 4a97c893071433d0d39cbf5261d0877f864c2189
149
+ - Foundation: https://github.com/FormalizedFormalLogic/Foundation, commit 54324e6e009f0d0a288897312d3feb1c0165ad19
150
+ - Mathlib: https://github.com/leanprover-community/mathlib4, tag v4.16.0
151
+ - Seymour: https://github.com/Ivan-Sergeyev/seymour, commit 27c0384977032b693daca8c4fcfc5cc274e1f2d6
152
+
153
+ miniCTX-v2 uses Lean version v4.16.0.
154
+
155
+ ## Citation
156
+
157
+ Please cite:
158
+ ```
159
+ @article{hu2024minictx,
160
+ title={miniCTX: Neural Theorem Proving with (Long-) Contexts},
161
+ author={Hu, Jiewen and Zhu, Thomas and Welleck, Sean},
162
+ journal={arXiv preprint arXiv:2408.03350},
163
+ year={2024}
164
+ }
165
+ ```