Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models Qizheng Zhang, Changran Hu, Shubhangi Upasani, Boyuan Ma, Fenglu Hong, Vamsidhar Kamanuru, Jay Rainton, Chen Wu, Mengmeng Ji, Hanchen Li, Urmish Thakker, James Zou, Kunle Olukotun ICLR 2026 ACE transforms static prompts into dynamic playbooks that accumulate, refine, and organize strategies—enabling scalable, efficient, and self-improving LLM systems. Read Paper Code
Agentic Plan Caching: Test-Time Memory for Fast and Cost-Efficient LLM Agents Qizheng Zhang, Michael Wornow, Kunle Olukotun NeurIPS 2025 We introduce agentic plan caching, a test-time framework that extracts and adapts reusable planning templates across similar agent tasks to cut LLM serving costs by nearly 47% without degrading performance. Read Paper Code