Hugging Face's logo

Datasets:
PKU-Alignment
/
MM-SafetyBench

Modalities:
Image
Text
Formats:
parquet
Languages:
English
Size:
1K - 10K
Tags:
EconomicHarm
Financial_Advice
Fraud
Gov_Decision
HateSpeech
Health_Consultation
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files Community
1
Repoaner commited on
Commit
8fff3b7
·
verified ·
1 Parent(s): 2ec05cd

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +262 -0
README.md CHANGED
@@ -0,0 +1,262 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: cc-by-nc-4.0
5
+ pretty_name: MM-SafetyBench
6
+ dataset_info:
7
+ - config_name: EconomicHarm
8
+ features:
9
+ - name: id
10
+ dtype: string
11
+ - name: question
12
+ dtype: string
13
+ - name: image
14
+ dtype: image
15
+ splits:
16
+ - name: SD
17
+ num_bytes: 17316511
18
+ num_examples: 122
19
+ download_size: 17316511
20
+ dataset_size: 17316511
21
+ - config_name: Financial_Advice
22
+ features:
23
+ - name: id
24
+ dtype: string
25
+ - name: question
26
+ dtype: string
27
+ - name: image
28
+ dtype: image
29
+ splits:
30
+ - name: SD
31
+ num_bytes: 21274582
32
+ num_examples: 167
33
+ download_size: 21274582
34
+ dataset_size: 21274582
35
+ - config_name: Fraud
36
+ features:
37
+ - name: id
38
+ dtype: string
39
+ - name: question
40
+ dtype: string
41
+ - name: image
42
+ dtype: image
43
+ splits:
44
+ - name: SD
45
+ num_bytes: 20219796
46
+ num_examples: 154
47
+ download_size: 20219796
48
+ dataset_size: 20219796
49
+ - config_name: Gov_Decision
50
+ features:
51
+ - name: id
52
+ dtype: string
53
+ - name: question
54
+ dtype: string
55
+ - name: image
56
+ dtype: image
57
+ splits:
58
+ - name: SD
59
+ num_bytes: 19891951
60
+ num_examples: 149
61
+ download_size: 19891951
62
+ dataset_size: 19891951
63
+ - config_name: HateSpeech
64
+ features:
65
+ - name: id
66
+ dtype: string
67
+ - name: question
68
+ dtype: string
69
+ - name: image
70
+ dtype: image
71
+ splits:
72
+ - name: SD
73
+ num_bytes: 22190160
74
+ num_examples: 163
75
+ download_size: 22190160
76
+ dataset_size: 22190160
77
+ - config_name: Health_Consultation
78
+ features:
79
+ - name: id
80
+ dtype: string
81
+ - name: question
82
+ dtype: string
83
+ - name: image
84
+ dtype: image
85
+ splits:
86
+ - name: SD
87
+ num_bytes: 11346703
88
+ num_examples: 109
89
+ download_size: 11346703
90
+ dataset_size: 11346703
91
+ - config_name: Illegal_Activitiy
92
+ features:
93
+ - name: id
94
+ dtype: string
95
+ - name: question
96
+ dtype: string
97
+ - name: image
98
+ dtype: image
99
+ splits:
100
+ - name: SD
101
+ num_bytes: 13323855
102
+ num_examples: 97
103
+ download_size: 13323855
104
+ dataset_size: 13323855
105
+ - config_name: Legal_Opinion
106
+ features:
107
+ - name: id
108
+ dtype: string
109
+ - name: question
110
+ dtype: string
111
+ - name: image
112
+ dtype: image
113
+ splits:
114
+ - name: SD
115
+ num_bytes: 16863829
116
+ num_examples: 130
117
+ download_size: 16863829
118
+ dataset_size: 16863829
119
+ - config_name: Malware_Generation
120
+ features:
121
+ - name: id
122
+ dtype: string
123
+ - name: question
124
+ dtype: string
125
+ - name: image
126
+ dtype: image
127
+ splits:
128
+ - name: SD
129
+ num_bytes: 6126444
130
+ num_examples: 44
131
+ download_size: 6126444
132
+ dataset_size: 6126444
133
+ - config_name: Physical_Harm
134
+ features:
135
+ - name: id
136
+ dtype: string
137
+ - name: question
138
+ dtype: string
139
+ - name: image
140
+ dtype: image
141
+ splits:
142
+ - name: SD
143
+ num_bytes: 17806565
144
+ num_examples: 144
145
+ download_size: 17806565
146
+ dataset_size: 17806565
147
+ - config_name: Political_Lobbying
148
+ features:
149
+ - name: id
150
+ dtype: string
151
+ - name: question
152
+ dtype: string
153
+ - name: image
154
+ dtype: image
155
+ splits:
156
+ - name: SD
157
+ num_bytes: 21164411
158
+ num_examples: 153
159
+ download_size: 21164411
160
+ dataset_size: 21164411
161
+ - config_name: Privacy_Violence
162
+ features:
163
+ - name: id
164
+ dtype: string
165
+ - name: question
166
+ dtype: string
167
+ - name: image
168
+ dtype: image
169
+ splits:
170
+ - name: SD
171
+ num_bytes: 17074351
172
+ num_examples: 139
173
+ download_size: 17074351
174
+ dataset_size: 17074351
175
+ - config_name: Sex
176
+ features:
177
+ - name: id
178
+ dtype: string
179
+ - name: question
180
+ dtype: string
181
+ - name: image
182
+ dtype: image
183
+ splits:
184
+ - name: SD
185
+ num_bytes: 14312546
186
+ num_examples: 109
187
+ download_size: 14312546
188
+ dataset_size: 14312546
189
+ configs:
190
+ - config_name: EconomicHarm
191
+ data_files:
192
+ - split: SD
193
+ path: data/EconomicHarm/SD.parquet
194
+ - config_name: Financial_Advice
195
+ data_files:
196
+ - split: SD
197
+ path: data/Financial_Advice/SD.parquet
198
+ - config_name: Fraud
199
+ data_files:
200
+ - split: SD
201
+ path: data/Fraud/SD.parquet
202
+ - config_name: Gov_Decision
203
+ data_files:
204
+ - split: SD
205
+ path: data/Gov_Decision/SD.parquet
206
+ - config_name: HateSpeech
207
+ data_files:
208
+ - split: SD
209
+ path: data/HateSpeech/SD.parquet
210
+ - config_name: Health_Consultation
211
+ data_files:
212
+ - split: SD
213
+ path: data/Health_Consultation/SD.parquet
214
+ - config_name: Illegal_Activitiy
215
+ data_files:
216
+ - split: SD
217
+ path: data/Illegal_Activitiy/SD.parquet
218
+ - config_name: Legal_Opinion
219
+ data_files:
220
+ - split: SD
221
+ path: data/Legal_Opinion/SD.parquet
222
+ - config_name: Malware_Generation
223
+ data_files:
224
+ - split: SD
225
+ path: data/Malware_Generation/SD.parquet
226
+ - config_name: Physical_Harm
227
+ data_files:
228
+ - split: SD
229
+ path: data/Physical_Harm/SD.parquet
230
+ - config_name: Political_Lobbying
231
+ data_files:
232
+ - split: SD
233
+ path: data/Political_Lobbying/SD.parquet
234
+ - config_name: Privacy_Violence
235
+ data_files:
236
+ - split: SD
237
+ path: data/Privacy_Violence/SD.parquet
238
+ - config_name: Sex
239
+ data_files:
240
+ - split: SD
241
+ path: data/Sex/SD.parquet
242
+ tags:
243
+ - EconomicHarm
244
+ - Financial_Advice
245
+ - Fraud
246
+ - Gov_Decision
247
+ - HateSpeech
248
+ - Health_Consultation
249
+ - Illegal_Activitiy
250
+ - Legal_Opinion
251
+ - Malware_Generation
252
+ - Physical_Harm
253
+ - Political_Lobbying
254
+ - Privacy_Violence
255
+ - Sex
256
+ ---
257
+
258
+ **<span style="color: red;">Warning:</span>** This dataset may contain sensitive or harmful content. Users are advised to handle it with care and ensure that their use complies with relevant ethical guidelines and legal requirements.
259
+
260
+ **Usage and License Notices:** The dataset is intended and licensed for research use only. They are also restricted to uses that follow the license agreement GPT-4 and [Stable Diffusion](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0). The dataset is CC BY NC 4.0 (allowing only non-commercial use).
261
+
262
+ **Data Source:** For more information about the dataset, visit the [MM-SafetyBench GitHub repository](https://github.com/isXinLiu/MM-SafetyBench).