To Err is AI : A Case Study Informing LLM Flaw Reporting Practices

Documenti analoghi

• WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs
  di: Han, Seungju, et al.
  Pubblicazione: (2024)
• Coordinated Flaw Disclosure for AI: Beyond Security Vulnerabilities
  di: Cattell, Sven, et al.
  Pubblicazione: (2024)
• WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models
  di: Jiang, Liwei, et al.
  Pubblicazione: (2024)
• What Makes it Ok to Set a Fire? Iterative Self-distillation of Contexts and Rationales for Disambiguating Defeasible Social and Moral Situations
  di: Rao, Kavel, et al.
  Pubblicazione: (2023)
• Malicious and Unintentional Disclosure Risks in Large Language Models for Code Generation
  di: Rabin, Rafiqul, et al.
  Pubblicazione: (2025)