Reproduce “ProMemAgent: Long-Horizon Proactive Assistance with Context-Aware Prospective Memory from Sensory Perceptions” 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
Lilin Xu, Linyang He, Kaiyuan Hou, Yuang Fan, Minghui Zhao, Xuhai Xu, Xiaofan Jiang
ProMemAgent links earlier implicit intentions to later situations where they may become actionable. It converts sensory events into text, stores facts with future cues and forward links, then retrieves relevant history for a reasoner to decide whether to respond. On CPMBench, Qwen3-8B reaches 75.2% Set-F1 versus GPT-5.4’s 65.5%, but this is a controlled benchmark, not continuous real-world assistance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.