The field of autonomous agents and large language models is rapidly evolving, with a focus on improving the efficiency, reliability, and performance of these systems. One of the key trends is the development of mechanisms to record, replay, and learn from experiences, allowing agents to adapt to new environments and improve their decision-making processes. Another area of research is the introduction of resource-aware multi-agent collaboration, which enables agents to work together more efficiently and effectively. Hierarchical memory structures and cross-task experiential learning are also being explored to enhance the capabilities of general agents. Furthermore, there is a growing interest in developing frameworks that enable large language models to autonomously transfer experiences from one task to another, reducing the need for extensive human labor. Notable papers in this area include: Get Experience from Practice: LLM Agents with Record & Replay, which proposes a new paradigm for agent development using record and replay mechanisms. Co-Saving: Resource Aware Multi-Agent Collaboration for Software Development, which introduces a resource-aware multi-agent system that leverages experiential knowledge to enhance operational efficiency. Efficiently Enhancing General Agents With Hierarchical-categorical Memory, which presents a general agent capable of learning without parameter updates using hierarchical memory retrieval and task-category oriented experience learning.