How and Why LLMs Generalize: A Fine-Grained Analysis of LLM Reasoning from Cognitive Behaviors to Low-Level Patterns

Documenti analoghi

• Climbing the Ladder of Reasoning: What LLMs Can-and Still Can't-Solve after SFT?
  di: Sun, Yiyou, et al.
  Pubblicazione: (2025)
• RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
  di: Sun, Yiyou, et al.
  Pubblicazione: (2025)
• Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations
  di: Sun, Yiyou, et al.
  Pubblicazione: (2025)
• The Long-Horizon Task Mirage? Diagnosing Where and Why Agentic Systems Break
  di: Wang, Xinyu Jessica, et al.
  Pubblicazione: (2026)
• Can Aha Moments Be Fake? Towards Quantifying Decorative and True Thinking in Chain-of-Thought
  di: Zhao, Jiachen, et al.
  Pubblicazione: (2025)