Add link to paper

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +6 -7
README.md CHANGED
@@ -1,23 +1,22 @@
1
  ---
2
- license: cc-by-4.0
3
  language:
4
  - en
5
- pretty_name: CAD Data
 
 
6
  task_categories:
7
  - text-classification
 
8
  tags:
9
  - prompt-injection
10
  - indirect-prompt-injection
11
  - agent-security
12
  - agentdojo
13
- size_categories:
14
- - 1K<n<10K
15
  ---
16
 
17
  # CAD Data
18
 
19
- Sentence-annotated indirect prompt injection data used to train and evaluate CAD, from
20
- *Robust Context-Aware Detection of Malicious Instructions in Text* (Liu, Ma, and Vorobeychik, 2026).
21
 
22
  Code: https://github.com/tavia-liu/CAD · Models: https://huggingface.co/tavialiu/CAD-models
23
 
@@ -88,4 +87,4 @@ using this data.
88
  journal = {arXiv preprint},
89
  year = {2026}
90
  }
91
- ```
 
1
  ---
 
2
  language:
3
  - en
4
+ license: cc-by-4.0
5
+ size_categories:
6
+ - 1K<n<10K
7
  task_categories:
8
  - text-classification
9
+ pretty_name: CAD Data
10
  tags:
11
  - prompt-injection
12
  - indirect-prompt-injection
13
  - agent-security
14
  - agentdojo
 
 
15
  ---
16
 
17
  # CAD Data
18
 
19
+ Sentence-annotated indirect prompt injection data used to train and evaluate CAD, from [Robust Context-Aware Detection of Malicious Instructions in Text](https://huggingface.co/papers/2608.05430) (Liu, Ma, and Vorobeychik, 2026).
 
20
 
21
  Code: https://github.com/tavia-liu/CAD · Models: https://huggingface.co/tavialiu/CAD-models
22
 
 
87
  journal = {arXiv preprint},
88
  year = {2026}
89
  }
90
+ ```