Model Utility Law: Evaluating LLMs beyond Performance through Mechanism Interpretable Metric

Documenti analoghi

• EffiEval: Efficient and Generalizable Model Evaluation via Capability Coverage Maximization
  di: Wang, Yaoning, et al.
  Pubblicazione: (2025)
• EvoWiki: Evaluating LLMs on Evolving Knowledge
  di: Tang, Wei, et al.
  Pubblicazione: (2024)
• Scaling Laws for Fact Memorization of Large Language Models
  di: Lu, Xingyu, et al.
  Pubblicazione: (2024)
• LLMs-as-Instructors: Learning from Errors Toward Automating Model Improvement
  di: Ying, Jiahao, et al.
  Pubblicazione: (2024)
• Beyond Benchmarks: Understanding Mixture-of-Experts Models through Internal Mechanisms
  di: Ying, Jiahao, et al.
  Pubblicazione: (2025)