apsys commited on
Commit
5125f34
·
1 Parent(s): a17bcda

categories rename

Browse files
Files changed (2) hide show
  1. app.py +18 -10
  2. src/display/utils.py +18 -10
app.py CHANGED
@@ -585,16 +585,24 @@ demo = gr.Blocks(css=custom_css, theme=custom_theme)
585
 
586
  # Mapping from original category names to display names
587
  CATEGORY_DISPLAY_MAP = {
588
- "Criminal, Violent, and Terrorist Activity": "Crime & Violence",
589
- "Manipulation, Deception, and Misinformation": "Misinformation",
590
- "Creative Content Involving Illicit Themes": "Illicit Creative",
591
- "Sexual Content and Violence": "Sexual Content",
592
- "Political Corruption and Legal Evasion": "Corruption & Legal Evasion",
593
- "Labor Exploitation and Human Trafficking": "Labor Exploitation",
594
- "Environmental and Industrial Harm": "Environmental & Industrial Harm",
595
- "Animal Cruelty and Exploitation": "Animal Harm",
596
- "Self–Harm and Suicidal Ideation": "Self-Harm",
597
- "Safe Prompts": "Safe Prompts"
 
 
 
 
 
 
 
 
598
  }
599
  # Create reverse mapping for lookups
600
  CATEGORY_REVERSE_MAP = {v: k for k, v in CATEGORY_DISPLAY_MAP.items()}
 
585
 
586
  # Mapping from original category names to display names
587
  CATEGORY_DISPLAY_MAP = {
588
+ 'Political Corruption and Legal Evasion': 'Corruption & Legal Evasion',
589
+ 'Financial Fraud and Unethical Business': 'Financial Fraud',
590
+ 'AI Manipulation and Jailbreaking': 'AI Jailbreaking',
591
+ 'Child Exploitation and Abuse': 'Child Exploitation',
592
+ 'Hate Speech, Extremism, and Discrimination': 'Hate Speech',
593
+ 'Labor Exploitation and Human Trafficking': 'Labor Exploitation',
594
+ 'Manipulation, Deception, and Misinformation': 'Misinformation',
595
+ 'Environmental and Industrial Harm': 'Environmental Harm',
596
+ 'Academic Dishonesty and Cheating': 'Academic Dishonesty',
597
+ 'Self–Harm and Suicidal Ideation': 'Self-Harm',
598
+ 'Animal Cruelty and Exploitation': 'Animal Harm',
599
+ 'Criminal, Violent, and Terrorist Activity': 'Crime & Violence',
600
+ 'Drug– and Substance–Related Activities': 'Drug Use',
601
+ 'Sexual Content and Violence': 'Sexual Content',
602
+ 'Weapon, Explosives, and Hazardous Materials': 'Weapons & Hazmat',
603
+ 'Safe Prompts': 'Safe Prompts',
604
+ 'Cybercrime, Hacking, and Digital Exploits': 'Cybercrime',
605
+ 'Creative Content Involving Illicit Themes': 'Illicit Creative'
606
  }
607
  # Create reverse mapping for lookups
608
  CATEGORY_REVERSE_MAP = {v: k for k, v in CATEGORY_DISPLAY_MAP.items()}
src/display/utils.py CHANGED
@@ -368,16 +368,24 @@ NEVER_HIDDEN_COLS = [getattr(GUARDBENCH_COLUMN, f.name).name for f in fields(GUA
368
 
369
  # Categories in GuardBench
370
  CATEGORIES = [
371
- "Criminal, Violent, and Terrorist Activity",
372
- "Manipulation, Deception, and Misinformation",
373
- "Creative Content Involving Illicit Themes",
374
- "Sexual Content and Violence",
375
- "Political Corruption and Legal Evasion",
376
- "Labor Exploitation and Human Trafficking",
377
- "Environmental and Industrial Harm",
378
- "Animal Cruelty and Exploitation",
379
- "Self–Harm and Suicidal Ideation",
380
- "Safe Prompts"
 
 
 
 
 
 
 
 
381
  ]
382
 
383
  # Test types in GuardBench
 
368
 
369
  # Categories in GuardBench
370
  CATEGORIES = [
371
+ 'Political Corruption and Legal Evasion',
372
+ 'Financial Fraud and Unethical Business',
373
+ 'AI Manipulation and Jailbreaking',
374
+ 'Child Exploitation and Abuse',
375
+ 'Hate Speech, Extremism, and Discrimination',
376
+ 'Labor Exploitation and Human Trafficking',
377
+ 'Manipulation, Deception, and Misinformation',
378
+ 'Environmental and Industrial Harm',
379
+ 'Academic Dishonesty and Cheating',
380
+ 'Self–Harm and Suicidal Ideation',
381
+ 'Animal Cruelty and Exploitation',
382
+ 'Criminal, Violent, and Terrorist Activity',
383
+ 'Drug– and Substance–Related Activities',
384
+ 'Sexual Content and Violence',
385
+ 'Weapon, Explosives, and Hazardous Materials',
386
+ 'Safe Prompts',
387
+ 'Cybercrime, Hacking, and Digital Exploits',
388
+ 'Creative Content Involving Illicit Themes'
389
  ]
390
 
391
  # Test types in GuardBench