Knowing What You Know Is Not Enough: Large Language Model Confidences Don't Align With Their Actions

Documenti analoghi

• Large Language Models Must Be Taught to Know What They Don't Know
  di: Kapoor, Sanyam, et al.
  Pubblicazione: (2024)
• vTune: Verifiable Fine-Tuning for LLMs Through Backdooring
  di: Zhang, Eva, et al.
  Pubblicazione: (2024)
• Dynamic Delayed Tree Expansion For Improved Multi-Path Speculative Decoding
  di: Thomas, Rahul, et al.
  Pubblicazione: (2026)
• Privacy-Preserving Mechanisms Enable Cheap Verifiable Inference of LLMs
  di: Pal, Arka, et al.
  Pubblicazione: (2026)
• Cascade: Token-Sharded Private LLM Inference
  di: Thomas, Rahul, et al.
  Pubblicazione: (2025)