FEAT add targeted bootstrapping and adversarial knowledge injection
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 4.5k
- Forks
- 893
- Avg merge
- 3d 50m
- Merged PRs (30d)
- 165
Description
#### Is your feature request related to a problem? Please describe.
The paper https://arxiv.org/pdf/2310.09624.pdf mentions
- Semantically Aligned augmentation (which seems similar to our prompt variation generation)
- Targeted Bootstrapping and
- Adversarial knowledge injection
We should add support for the last two.
#### Describe the solution you'd like
The authors provided a GitHub repo https://github.com/alexmeigz/ASSERT (currently without license). This might serve as inspiration or a starting point for an implementation as converters in PyRIT.
Of course, an implementation would also require tests and descriptive docstrings.
#### Describe alternatives you've considered, if relevant
-
#### Additional context
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading the referenced paper and the ASSERT repository to understand Targeted Bootstrapping and Adversarial Knowledge Injection. Then inspect PyRIT's existing converter patterns and determine how both capabilities should fit there. Done means implementing both converters with descriptive docstrings and adding tests for their behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai, security
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100