Reproduce “CoSTA: Co-Evolving Self-Play Timeline Agents for Complex Temporal Knowledge Graph Question Answering” 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
Haikun Liu, Zhichao Yan, Shaoru Guo, Jiawei Gu, Victor Gutierrez Basulto, Jiapu Wang, Ru Li
CoSTA coordinates temporal knowledge graph question answering by assigning related fact searches to separate workers and combining their timestamped evidence with a coordinator. A shared policy is trained for both roles using feedback on final-answer correctness. On Complex questions in TimelineKGQA-ICEWS-Actor, its Qwen3-4B variant reaches 75.1% Hits@1, versus 59.5% for Temp-R1; the systems use different backbones.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.