Reproduce “Introduction to Automated Negotiation” 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

Dave de Jonge

*Introduction to Automated Negotiation* is an open-access instructional manuscript that builds a practical and formal foundation for autonomous agents negotiating over shared outcomes. It combines mathematical models, negotiation strategies, opponent modeling, game-theoretic concepts, empirical evaluation methods, multilateral protocols, and Python-based exercises through the accompanying NegoSim simulator.

Key insights

  • Negotiation quality requires balancing agreement probability with the value of an agreement; agreement rate or utility alone can give misleading conclusions about agent performance.
  • Opponent modeling can help, but simple strategies such as MiCRO may remain competitive on established benchmarks, showing that benchmark structure can favor relatively lightweight methods.
  • Theoretical bargaining solutions rely on restrictive assumptions, so empirical evaluation should include self-play, diverse opponents and domains, uncertainty estimates, statistical tests, and complementary ranking procedures.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.