From Hallucination to Scheming: A Unified Taxonomy and Benchmark Analysis for LLM Deception

Documenti analoghi

• Cheap Talk, Empty Promise: Frontier LLMs easily break public promises for self-interest
  di: Shi, Jerick, et al.
  Pubblicazione: (2026)
• CoopEval: Benchmarking Cooperation-Sustaining Mechanisms and LLM Agents in Social Dilemmas
  di: Tewolde, Emanuel, et al.
  Pubblicazione: (2026)
• Domain-Independent Deception: A New Taxonomy and Linguistic Analysis
  di: Verma, Rakesh M., et al.
  Pubblicazione: (2024)
• Causality for Natural Language Processing
  di: Jin, Zhijing
  Pubblicazione: (2025)
• Now, Later, and Lasting: Ten Priorities for AI Research, Policy, and Practice
  di: Horvitz, Eric, et al.
  Pubblicazione: (2024)