arXiv 2601.14192
Toward Efficient Agents: Memory, Tool learning, and Planning
By Xiaofang Yang, Lijun Li, et al.
Published 2026-01-20
Wiki summary
Explore the paper's summary, context, and related research on Papiers.
Recent years have witnessed increasing interest in extending large language models into agentic systems. While the effectiveness of agents has continued to improve, efficiency, which is crucial for real-world deployment, has often been overlooked. This paper therefore investigates efficiency from three core components of agents: memory, tool learning, and planning, considering costs such as latency, tokens, steps, e…