Reproduce “Auto-SLURP: A Benchmark Dataset for Evaluating Multi-Agent Frameworks in Smart Personal Assistant” 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
Lei Shen, Xiaoyu Shen
Auto-SLURP introduces a benchmark dataset for evaluating multi-agent frameworks in smart personal assistants, combining 1,100 labeled queries with simulated execution servers across 23 domains while revealing significant performance gaps between leading frameworks like AgentLite, AutoGen, and CamelAI in end-to-end task completion.
Key insights
- Intent recognition remains a major bottleneck across all frameworks
- Framework design choices significantly impact success rates (AgentLite's "think and react" vs LangGraph's simpler prompting)
- Finetuning individual components like intent agents can provide substantial improvements
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.