Do Not Design, Learn: A Trainable Scoring Function for Uncertainty Estimation in Generative LLMs

Documenti analoghi

• MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs
  di: Bakman, Yavuz Faruk, et al.
  Pubblicazione: (2024)
• Uncertainty Quantification for Hallucination Detection in Large Language Models: Foundations, Methodology, and Future Directions
  di: Kang, Sungmin, et al.
  Pubblicazione: (2025)
• Reconsidering LLM Uncertainty Estimation Methods in the Wild
  di: Bakman, Yavuz, et al.
  Pubblicazione: (2025)
• CroMo-Mixup: Augmenting Cross-Model Representations for Continual Self-Supervised Learning
  di: Mushtaq, Erum, et al.
  Pubblicazione: (2024)
• Hair-Trigger Alignment: Black-Box Evaluation Cannot Guarantee Post-Update Alignment
  di: Bakman, Yavuz, et al.
  Pubblicazione: (2026)