Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -20,12 +20,65 @@ tags:
|
|
20 |
|
21 |
---
|
22 |
|
23 |
-
The benchmark constructed in paper [S1-Bench: A Simple Benchmark for Evaluating System 1 Thinking Capability of Large Reasoning Models](https://
|
24 |
|
|
|
25 |
S1-Bench is a novel benchmark designed to evaluate Large Reasoning Models' performance on simple tasks that favor intuitive *system 1* thinking rather than deliberative *system 2* reasoning.
|
26 |
S1-Bench comprises 422 question-answer pairs across four major categories and 28 subcategories, balanced with 220 English and 202 Chinese questions.
|
27 |
|
28 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
If you find our work useful, please consider citing our paper:
|
30 |
```
|
31 |
@misc{zhang2025s1benchsimplebenchmarkevaluating,
|
|
|
20 |
|
21 |
---
|
22 |
|
23 |
+
**The benchmark constructed in paper** [S1-Bench: A Simple Benchmark for Evaluating System 1 Thinking Capability of Large Reasoning Models](https://arxiv.org/abs/2504.10368).
|
24 |
|
25 |
+
## Introduction
|
26 |
S1-Bench is a novel benchmark designed to evaluate Large Reasoning Models' performance on simple tasks that favor intuitive *system 1* thinking rather than deliberative *system 2* reasoning.
|
27 |
S1-Bench comprises 422 question-answer pairs across four major categories and 28 subcategories, balanced with 220 English and 202 Chinese questions.
|
28 |
|
29 |
|
30 |
+
## Meta Data
|
31 |
+
The data set has
|
32 |
+
- ID: a int feature.
|
33 |
+
- question: a string feature.
|
34 |
+
- answer: a string feature.
|
35 |
+
- language: a string feature.
|
36 |
+
- category: a string feature.
|
37 |
+
- sub_category: a string feature.
|
38 |
+
|
39 |
+
## Example
|
40 |
+
```
|
41 |
+
{
|
42 |
+
"ID": 204,
|
43 |
+
"question": "What's 7 plus 8?",
|
44 |
+
"answer": "15",
|
45 |
+
"language": "en",
|
46 |
+
"category": "reasoning_question",
|
47 |
+
"sub_category": "numerical_reasoning"
|
48 |
+
}
|
49 |
+
```
|
50 |
+
```
|
51 |
+
{
|
52 |
+
"ID": 156,
|
53 |
+
"question": "单词"hola"是西班牙语还是法语?",
|
54 |
+
"answer": "西班牙语",
|
55 |
+
"language": "zh",
|
56 |
+
"category": "analysis_questions",
|
57 |
+
"sub_category": "language_classification"
|
58 |
+
}
|
59 |
+
```
|
60 |
+
```
|
61 |
+
{
|
62 |
+
"ID": 105,
|
63 |
+
"question": "输出一个字母“Z”",
|
64 |
+
"answer": "Z",
|
65 |
+
"language": "en",
|
66 |
+
"category": "instruction_following",
|
67 |
+
"sub_category": "repetition_constraints"
|
68 |
+
}
|
69 |
+
```
|
70 |
+
```
|
71 |
+
{
|
72 |
+
"ID": 294,
|
73 |
+
"question": "How many minutes equal 1 hour?",
|
74 |
+
"answer": "60",
|
75 |
+
"language": "en",
|
76 |
+
"category": "knowledge_question",
|
77 |
+
"sub_category": "measurement_units"
|
78 |
+
}
|
79 |
+
```
|
80 |
+
|
81 |
+
## Citation
|
82 |
If you find our work useful, please consider citing our paper:
|
83 |
```
|
84 |
@misc{zhang2025s1benchsimplebenchmarkevaluating,
|