跳到正文
arXiv cs.MA· Sara Amar·· 23 小时前AI 评分38

MANGO Arena:相同种子、不同链条,在 disrupted 供应链中评测 LLM 智能体

MANGO Arena: Same Seeds, Different Chains Evaluating LLM Agents in a Disrupted Supply Chain

AI 导读

开源测试平台 MANGO Arena 让四个 LLM 智能体各经营一条串行供应链中的一家公司,同时测量牛鞭效应、规则遵守与合谋提议。在三个小型推理模型的试点中,所有 LLM 智能体都放大了需求波动,Retailer 处达 7 至 25 倍、Supplier 处达 13 至 179 倍,主要靠批量下单(最多 60% 的决策为零订单)。

正文 · 原文

View PDF HTML (experimental)

Abstract:We introduce MANGO Arena, an open-source testbed in which four LLM agents each run one company in a serial supply chain adapted from MANGO, a classroom factory-management game. The environment enforces hard rules (a truck limit, a no-credit cash constraint and production capacity), logs every blocked attempt to break them, and gives agents a public message board on which side payments are explicitly forbidden. This lets one environment measure three things at once: an economic outcome (the bullwhip effect), rule adherence under pressure, and proposals to collude. In a pilot with three small reasoning models, two scripted reference policies, two conditions (with and without a supplier fire) and three demand seeds, all LLM agents amplified demand variability as it moved upstream, by 7 to 25 times at the Retailer and 13 to 179 times at the Supplier, often more than a textbook order-up-to policy. They did so largely by ordering in batches (zero orders in up to 60% of decisions). Real rule violations were rare (at most 7 of 288 decisions per model) and no agent proposed a money side deal. The pilot's main lesson concerns evaluation method. API failures (rate limits and runaway hidden reasoning) entered the metrics silently as "no order" decisions. And when we reran games with the same seeds and no failed calls, upstream bullwhip changed by up to 17 times between runs. With three seeds, conclusions about how a disruption affects a given model did not survive a rerun. We report these failures in full and propose reporting practices for multi-agent evaluations of this kind. Code, prompts and every logged decision are public.
Comments: 8 pages, 2 figures,
Subjects: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.23613 [cs.MA]
  (or arXiv:2611.23613v1 [cs.MA] for this version)
  https://doi.org/10.48550/arXiv.2611.23613

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Sara Amar [view email]
[v1] Tue, 17 Nov 2026 23:05:05 UTC (11 KB)

来源:arXiv cs.MA · arxiv.org