Reproduce “Knowing Whom to Trust is not Enough: The Trust--Agreement Gap in Repeated Multi-Party LLM 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
Amit Sarker, Moumita Choudhury, Saaduddin Mahmud, Shlomo Zilberstein
TASMAN carries directed trust scores between negotiations and tests whether prompted agents use them when proposing and voting. In a Grok-4.1-Fast reasoning-agent setting, using the same BRS trust with prescriptive guidance raised agreement from 35.5% to 66.0%; full prescriptive guidance reached 68.5%, 23 points above no trust. Findings are limited to the studied settings and do not establish efficiency gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.