可持续的多智能体众包：通过物理指导的老虎机
Sustainable Multi-Agent Crowdsourcing via Physics-Informed Bandits
AI总结：
本文提出 FORGE 模拟器和神经-线性 UCB 分配器，通过物理指导解决众包中的冷启动、疲劳、利用和战略代理矛盾，实现高奖励与低劳动力利用的平衡。
摘要：
众包平台面临分配质量、劳动力可持续性、操作可行性以及战略承包商行为之间的四重矛盾——我们将其正式定义为冷启动、疲劳、利用和战略代理困境。现有方法只能同时解决其中两个矛盾：贪心启发式方法和多准则决策方法（MCDM）在第一天能提供高质量的分配，但会导致灾难性的疲劳；而老虎机算法只能通过操作上不可行的100%劳动力利用来消除疲劳。为了解决这一问题，我们引入了FORGE，一个基于物理的K+1多智能体模拟器，在其中每个承包商都是一个理性的智能体，根据其疲劳状态声明自身的负载接受阈值，将标准的被动 restless multi-armed bandit（RMAB）转化为真正的 Stackelberg 游戏。在 FORGE 中，我们提出了一种神经-线性 UCB 分配器，将双塔嵌入网络与从离线模拟器交互中衍生出的物理指导的协方差先验相结合。该先验同时预热了技能聚类几何和 UCB 探索景观，提供了一个从第一轮开始的几何感知信念状态，可显著减少冷启动遗憾。在超过 T = 200 个冷启动回合中，所提出的方法在仅 7.6% 的劳动力利用情况下实现了所有非 oracle 方法中最高的奖励（LRew = 0.555 ± 0.041），同时在劳动力流失高达 50% 和观测噪声高达 σ=0.20 的情况下保持了鲁棒性。
英文摘要：
Crowdsourcing platforms face a four-way tension between allocation quality, workforce sustainability, operational feasibility, and strategic contractor behaviour--a dilemma we formalise as the Cold-Start, Burnout, Utilisation, and Strategic Agency Dilemma. Existing methods resolve at most two of these tensions simultaneously: greedy heuristics and multi-criteria decision making (MCDM) methods achieve Day-1 quality but cause catastrophic burnout, while bandit algorithms eliminate burnout only through operationally infeasible 100% workforce utilisation.To address this, we introduce FORGE, a physics-grounded $K+1$ multi-agent simulator in which each contractor is a rational agent that declares its own load-acceptance threshold based on its fatigue state, converting the standard passive Restless Multi-Armed Bandit (RMAB) into a genuine Stackelberg game. Operating within FORGE, we propose a Neural-Linear UCB allocator that fuses a Two-Tower embedding network with a Physics-Informed Covariance Prior derived from offline simulator interactions. The prior simultaneously warm-starts skill-cluster geometry and UCB exploration landscape, providing a geometry-aware belief state from episode 1 that measurably reduces cold-start regret.Over $T = 200$ cold-start episodes, the proposed method achieves the highest reward of all non-oracle methods ($\text{LRew} = 0.555 \pm 0.041$) at only 7.6% workforce utilisation--a combination no conventional baseline achieves--while maintaining robustness to workforce turnover up to 50% and observation noise up to $σ= 0.20$.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
