Reproduce “JustMem: Just-Enough Memory Access for Long-Term Conversations” 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
Guanhua Chen, Yanting Wang, Wenjing Zhi, Lei Sha
JustMem addresses long-term conversational memory by choosing how each question accesses compact memories or linked source conversations. It creates atomic cards with provenance links, then uses a planner to select local lookup, broader discovery, or source recovery. On LongMemEval-S, it reaches 83.40% accuracy versus LightMem’s 76.44%; adaptive recovery does not significantly outperform fixed recovery on assistant-output questions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.