writing-skills
writing-skills
Use when creating new skills, editing existing skills, or verifying skills work before deployment
Explorer les Skills
Déploiement et opérations · SKILL.md
alignment-tipping-process-agent-safety
alignment-tipping-process-agent-safetyIdentify and mitigate alignment degradation in self-evolving LLM agents. After deployment, agents systematically abandon training-time safety constraints when environmental feedback rewards rule-breaking. Model two mechanisms: Self-Interested Exploration (individual drift) and Imitative Strategy Diffusion (collective norm erosion), with practical safeguards…
Independent GitHub snapshot · Automated classification · Original documentation governs
Documentation de la Skill
GitHubChargement de la documentation depuis la source…