Large language models struggle with cumulative state tracking across multiple updates despite strong single-step arithmetic, and their measured acc...
论文概览
- arXiv ID:
2603.27343 - 发布机构 / 来源:Hugging Face Daily Papers
- 作者:Dengzhe Hou, Lingyu Jiang, Deng Li et al.
- 社区热度:3 Upvotes
核心摘要(Abstract / AI Summary)
Large language models struggle with cumulative state tracking across multiple updates despite strong single-step arithmetic, and their measured accuracy depends heavily on output protocols.
