Reproduce “A Self-Exploring and Self-Evolving Paradigm for Adaptive Edge AI Development” 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
Leming Shen, Qiang Yang, Xinyu Huang, Zijing Ma, Yuanqing Zheng, Chris Xiaoxuan Lu
SE² lets a coordinator choose which specialist agents retrieve information, reason about context, write Python code, or test next. Device-side probes gather runtime feedback, while offline reinforcement fine-tuning uses feedback and graders. On XRF55 human-activity recognition, deployed on a Raspberry Pi with 8 GB RAM and no GPU, the authors report 26.5% higher classification accuracy than baselines, limited to the tested setting.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.