Anchor Points: Benchmarking Models with Much Fewer Examples

Documenti analoghi

• KTO: Model Alignment as Prospect Theoretic Optimization
  di: Ethayarajh, Kawin, et al.
  Pubblicazione: (2024)
• Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token Budgets
  di: Tran, Dat, et al.
  Pubblicazione: (2026)
• Understanding Dataset Difficulty with $\mathcal{V}$-Usable Information
  di: Ethayarajh, Kawin, et al.
  Pubblicazione: (2021)
• Data Checklist: On Unit-Testing Datasets with Usable Information
  di: Zhang, Heidi C., et al.
  Pubblicazione: (2024)
• Document Optimization for Black-Box Retrieval via Reinforcement Learning
  di: Uzan, Omri, et al.
  Pubblicazione: (2026)