#artificial intelligence
May 2026
From History to State: Constant-Context Skill Learning for LLM Agents
constant-context skill learning is proposed, a context-to-weights framework for recurring agent workflows: reusable procedures are learned in lightweight task-family modules, while inference conditions only on the current observation and a compact state block.
Hao-Yang Xie, X. Wang, Yan-Cheng Wang et al.
· arXiv.org · 6 citations