← 返回论文库

Generative Agents: Interactive Simulacra of Human Behavior

Park, et al. (Stanford + Google) · 2023
L3.4 · LLM Agent Patterns & FrameworksUIST 2023#memory#simulation
CORE IDEA
25 个 LLM agent 在模拟小镇：每个有 memory stream + reflection + planning，emergent 社交行为。
L-ANCHOR · 为什么在这一层重要
agent simulation + memory
arXiv:2304.03442 ↗源码 ↗

相关论文

FinMem: A Performance-Enhanced LLM Trading Agent with Layered Memory
L0.22023
FactorMiner: Self-Evolving Alpha Discovery with Memory
L0.52025
MemGPT: Towards LLMs as Operating Systems
L3.42023
Voyager: An Open-Ended Embodied Agent with Large Language Models
L3.42023