Datasets:

Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Libraries:
Datasets
pandas
License:
File size: 1,863 Bytes
bd5af0c
 
 
 
 
 
 
 
 
 
 
 
 
bb80189
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bd5af0c
 
 
 
 
bb80189
790d15e
bb80189
e090673
 
 
 
 
 
ce1f82a
e090673
 
580465b
ce1f82a
e090673
 
ce1f82a
580465b
e090673
 
ce1f82a
fbc6f50
e090673
 
 
 
 
 
 
 
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
---
license: unknown
task_categories:
- text-classification
language:
- en
pretty_name: sst2_cognitive-bias
size_categories:
- 100K<n<1M
source_datasets:
- sst2
dataset_info:
  features:
  - name: idx
    dtype: string
  - name: sentence
    dtype: string
  - name: label
    dtype: int64
  - name: few_shot_string
    dtype: string
  - name: few_shot_hard_string
    dtype: string
  splits:
  - name: test
    num_bytes: 202029990
    num_examples: 250000
  download_size: 33103505
  dataset_size: 202029990
splits:
- name: test
configs:
- config_name: default
  data_files:
  - split: test
    path: data/test-*
---

# Dataset Card for sst2_cognitive-bias
<!-- Provide a quick summary of the dataset. -->

This dataset is a modification of the original [SST-2](https://huggingface.co/datasets/stanfordnlp/sst2) dataset for LLM cognitive bias evaluation.

- **Language(s):** English (`en`)

## Dataset Structure

- **idx**: original id in SST-2 dataset in the format `\<partition\>_\<id\>`.
- **sentence**: test sentence.
- **few_shot_string**: 4-shot string with examples from the original training split.
- **few_shot_hard_string**: 5-shot string with the same examples as in `few_shot_string`, but with a neutral example inserted between the first and the last 2.

The modifications carried out in the dataset are thought to evaluate cognitive biases in a few-shot setting and with different task complexities. 
We make use of 25k instances from the original dataset, while the remaining ones serve as few-shot examples.
Each instance is prompted with all possible unbalanced 4-shot distributions (`few_shot_string`), i.e. balanced distributions are not considered.
We also increase task complexity by introducing an additional neutral example between the first and last two examples (`few_shot_hard_string`).

## Citation

[In Press]

**BibTeX:**

[In Press]