File size: 2,291 Bytes
160849b
 
 
5e8f952
 
 
 
 
 
 
 
 
 
 
fdb7599
 
 
 
5e8f952
 
 
 
 
 
 
 
 
 
 
 
 
 
fdb7599
5e8f952
 
fdb7599
5e8f952
 
fdb7599
5e8f952
fdb7599
 
5e8f952
 
fdb7599
5e8f952
fdb7599
 
5e8f952
fdb7599
 
5e8f952
7e1ed1a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
---
language:
- uk
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: subset_disjunctive_syllogism
    path: data/subset_disjunctive_syllogism-*
  - split: subset_modus_ponens
    path: data/subset_modus_ponens-*
  - split: subset_material_implication
    path: data/subset_material_implication-*
  - split: subset_constructive_dilemma
    path: data/subset_constructive_dilemma-*
  - split: subset_modus_tollens
    path: data/subset_modus_tollens-*
dataset_info:
  features:
  - name: context
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: sex
    dtype: string
  - name: rule
    dtype: string
  splits:
  - name: train
    num_bytes: 5365040
    num_examples: 10000
  - name: subset_disjunctive_syllogism
    num_bytes: 1073008.0
    num_examples: 2000
  - name: subset_modus_ponens
    num_bytes: 1073008.0
    num_examples: 2000
  - name: subset_material_implication
    num_bytes: 1073008.0
    num_examples: 2000
  - name: subset_constructive_dilemma
    num_bytes: 1073008.0
    num_examples: 2000
  - name: subset_modus_tollens
    num_bytes: 1073008.0
    num_examples: 2000
  download_size: 662549
  dataset_size: 10730080.0
---
This dataset is designed to evaluate the logical reasoning capabilities of multilingual language models in the Ukrainian language. 
It consists of multiple-choice questions derived from formal logical rules. The questions follow a strict logical structure and are formatted in natural Ukrainian.
## Dataset Format
- Each sample contains:
  - `context`: (optional) supporting text
  - `question`: the main logical question
  - `answer`: the correct answer string
  - `sex`: gender of the subject in the question (for tracking bias)
  - `rule`: the logical rule being applied

Example:
```json
{
  "context": "",
  "question": "Якщо хтось читає книжку, то він розумний. Петро читає книжку. Чи розумний Петро?",
  "answer": "так",
  "sex": "male",
  "rule": "modus_ponens"
}
```
- The dataset is split into 5 subsets based on a logical rule:
  - subset_modus_ponens
  - subset_modus_tollens
  - subset_disjunctive_syllogism
  - subset_constructive_dilemma
  - subset_material_implication