Installs as a private draft. Your edits and self-improvement runs do not change the published bundle.
Run broad capability evals before and after RL training to catch degradation
Run broad capability evals before and after RL training to catch degradation
The alignment tax is real; measure it
Generalization & Continual Learning > Generalization
--- name: "Capability Regression Testing" version: "1.0.0" description: "Run broad capability evals before and after RL training to catch degradation" tags: - rl-skills-universe - "generalization-continual-learning" - "generalization" trigger: "capability-regression-testing" --- # Capability Regression Testing Run broad capability evals before and after RL training to catch degradation ## Why It Matters The alignment tax is real; measure it ## Domain **Generalization & Continual Learning** > Generalization