Technical working session
Talk to a researcher about your post-training needs
What you leave with
Five runnable tasks in Harbor format to evaluate in your own environment after the session.
Meet with a researcher building RL gyms and expert-verified data for frontier-model post-training. We’ll review your model, harness, target capability, and the failure surface you want to improve.
In the session
- Review the capability and failure surface you want to improve
- Examine how it appears in your current harness and evaluations
- Map it to a runnable task, environment, or data strategy
Prefer email?
Email info@collinear.ai