When Parts Are Greater Than Sums: Individual LLM Components Can Outperform Full Models

Documenti analoghi

• Do Localization Methods Actually Localize Memorized Data in LLMs? A Tale of Two Benchmarks
  di: Chang, Ting-Yun, et al.
  Pubblicazione: (2023)
• Language Models can Infer Action Semantics for Symbolic Planners from Environment Feedback
  di: Zhu, Wang, et al.
  Pubblicazione: (2024)
• "The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework
  di: Cui, Jin, et al.
  Pubblicazione: (2026)
• PDDL-Mind: Large Language Models are Capable on Belief Reasoning with Reliable State Tracking
  di: Zhu, Wang Bill, et al.
  Pubblicazione: (2026)
• PSALM-V: Automating Symbolic Planning in Interactive Visual Environments with Large Language Models
  di: Zhu, Wang Bill, et al.
  Pubblicazione: (2025)