← back to paper
arxiv: 2605.09823 · 2 revisions
CalBench: Evaluating Coordination-Privacy Trade-offs in Multi-Agent LLMs