Reproduce “FlowState: Execution State as Memory for Long-Horizon LLM Agents” 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
Minghao Li, Bangyan Li, Zifan Wang, Yulong Li, Hu Xu, Gan Zhang, Jingtong Wu, Wenqiang Xu
FlowState uses task state as working and cross-request memory. It incrementally updates state nodes outside the prompt and progressively discloses historical details when needed. Against Full Context with DeepSeek-V4-Flash, it raises success by 4.55 percentage points across four MemoryArena environments and pass rate by 13.95 points across two τ³-Bench domains. These results do not establish safety.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.