LAION-AI / LAION-AI/Open-Assistant
Assistant Output Veracity Checkbox
Open
Nobody has claimed this yet.
data
- Dominant language
- Python
- Stars
- 37.4k
- Forks
- 3.3k
- PR merge metrics
- No merged PRs in 30d
Description
We should add some way of indicating that synthetic output is false. We really should be using such data to train the reward model I imagine?
Contributor guide
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 locating where assistant outputs are recorded and where data is prepared for reward-model training. Define how false synthetic output should be indicated and how that signal reaches the training data; the work is done when the behavior and resulting data are clearly specified and covered by appropriate validation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100