BELLS: A Framework Towards Future Proof Benchmarks for the Evaluation of LLM Safeguards

Documenti analoghi

• The bitter lesson of misuse detection
  di: Mariaccia, Hadrien, et al.
  Pubblicazione: (2025)
• MAGE: Safeguarding LLM Agents against Long-Horizon Threats via Shadow Memory
  di: Wang, Yuhui, et al.
  Pubblicazione: (2026)
• Covert Malicious Finetuning: Challenges in Safeguarding LLM Adaptation
  di: Halawi, Danny, et al.
  Pubblicazione: (2024)
• Towards Reliable and Practical LLM Security Evaluations via Bayesian Modelling
  di: Llewellyn, Mary, et al.
  Pubblicazione: (2025)
• Protecting Users From Themselves: Safeguarding Contextual Privacy in Interactions with Conversational Agents
  di: Ngong, Ivoline, et al.
  Pubblicazione: (2025)