Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination

About

Multi-agent large language model (LLM) systems often fail to reliably outperform a single strong model equipped with best-of-N sampling. We argue that a core source of this instability is ill-posed equilibrium selection: current systems specify what information agents share, but not which coordination convention should be selected. We formalize a broad class of such systems as discounted incomplete-information Markov games and show that two common pathologies, oscillation between competing conventions and drift across them, can both induce unstable learning and linear Bayesian regret. To obtain a well-posed target, we introduce the Heterogeneous Quantal Response Equilibrium (HQRE), an entropy-regularized equilibrium concept with agent- and state-dependent temperatures. Under a monotonicity condition, HQRE is unique, admits linearly convergent mirror updates, and yields bounded Bayesian regret; the same condition yields rollout-measurable stability diagnostics. We instantiate this objective in two algorithms: DICE-PC, which coordinates frozen models through prompt-control actions, and DICE-FT, which performs parameter-efficient mirror fine-tuning. Across eleven benchmarks in four domains, DICE improves accuracy-cost trade-offs over strong within-class baselines; on reasoning and planning tasks, DICE-PC improves by 4.3 percentage points on average and DICE-FT by 8.5 points.

Yi Xie, Zhanke Zhou, Chentao Cao, Bo Liu, Bo Han• 2026

Related benchmarks

TaskDatasetResultRank
Mathematical ReasoningMATH 500
Accuracy99.1
589
Logic reasoningZebraLogic
Accuracy93.1
49
Travel PlanningTravelPlanner (test)
Final Pass Rate15.2
49
PlanningPlanBench
Accuracy42.7
47
Travel PlanningTravelPlanner (val)
Delivery Rate100
42
Mathematical ReasoningAIME24
Accuracy (AIME24)86.3
34
PlanningAutoLogic
Accuracy89.1
34
Multi-agent coordinationOvercooked-AI Mean of CR/AA/Ring/FC/CC layouts
Self-Play Score121.6
16
Multi-agent coordinationCollabEscape (test)
Success Rate80
13
Multi-agent coordinationCollabCapture (test)
Success Rate100
13
Showing 10 of 10 rows

Other info

Follow for update