Reproduce “Generalized Multi-agent Social Simulation Framework” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Gang Li, Jie Lin, Yining Tang, Ziteng Wang, Yirui Huang, Junyu Zhang, Shuang Luo, Yike Guo and 1 more
The authors present a modular framework for social simulations and a method for summarizing retrieved memories. It combines reusable worlds, agents, memory, workflows, and behaviors; the method identifies questions and extracts details. In an example, agents connect events. On RoleAgentBench, memory-system win rate rises from 25.60 without summaries to 43.90 with them; this benchmark result does not establish real-world behavior.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.