智能体
arxiv.org ·
递归推理还是统计外推？多智能体相互依赖决策中的上下文学习
该 arXiv 论文研究 LLM 智能体在多智能体不完全信息博弈中的上下文学习（ICL）机制，区分其收益来自内部推理还是统计模式外推。作者构建公共品博弈并操纵历史反馈的统计结构，以与历史无关的理性预期均衡（REE）为基准评估决策质量。实验显示，当历史统计模式被破坏时，更长上下文的收益大幅消失，决策质量退化至无上下文基线，且该效应随战略相互依赖增强而放大，表明此类战略环境中的 ICL 行为更符合统计外推而非策略推理。论文将 ICL 机制研究扩展到战略多智能体场景，并提出 REE 作为区分推理与外推的诊断工具。
AI推荐理由 arXiv 预印本，作者与摘要明确，实验设计（公共品博弈、操纵历史反馈统计结构、以 REE 为基准）可验证，信息密度高。核心新信息是：在多智能体不完全信息博弈中，当历史统计模式被破坏后，更长上下文的收益基本消失，决策质量退化到无上下文基线，且战略相互依赖越强退化越明显，据此认为 ICL 更接近统计外推而非递归信念推理。局限是仅基于摘要，未见具体数据与规模，结论限于该类战略环境，不能外推为 LLM 普遍不具备推理能力。
78
正文
Computer Science > Artificial Intelligence
arXiv:2609.18591 (cs)
[Submitted on 16 Sep 2026]
Title:Recursive Reasoning or Statistical Extrapolation? In-Context Learning in Multi-Agent Interdependent Decision-Making
Authors:Yu Liu, Wenwen Li, Yifan Dou, Guangnan Ye
View a PDF of the paper titled Recursive Reasoning or Statistical Extrapolation? In-Context Learning in Multi-Agent Interdependent Decision-Making, by Yu Liu and 3 other authors
View PDF
HTML (experimental)
Abstract:In-context learning (ICL) enables large language model (LLM) agents to improve decisions using interaction history, yet it remains unclear whether such improvement reflects refined internal reasoning or mere extrapolation of statistical patterns. To disentangle these mechanisms, we study LLM agents in multi-agent incomplete-information games that require recursive belief reasoning. By constructing a public goods game and manipulating the statistical structure of historical feedback, we evaluate decision quality against a history-independent rational expectations equilibrium (REE) benchmark. Our experiments reveal that when historical statistical patterns are disrupted, the benefits of longer context largely vanish, degrading decision quality to the no-context baseline in a way sharply amplified by stronger strategic interdependence. These results suggest that, in such strategic environments, ICL behavior is more consistent with statistical extrapolation than with strategic reasoning. Our work extends the mechanistic study of ICL to strategic multi-agent settings, introduces REE as a diagnostic tool for distinguishing reasoning from extrapolation, and provides a reusable framework for probing the boundaries of LLM reasoning in recursive belief tasks.
Subjects:
Artificial Intelligence (cs.AI); General Economics (econ.GN)
Cite as:
arXiv:2609.18591 [cs.AI]
(or
arXiv:2609.18591v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.18591
Focus to learn more
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Yu Liu [view email]
[v1]
Wed, 16 Sep 2026 12:49:19 UTC (9,400 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Recursive Reasoning or Statistical Extrapolation? In-Context Learning in Multi-Agent Interdependent Decision-Making, by Yu Liu and 3 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.AI
new
|
recent
| 2026-09
Change to browse by:
cs
econ
econ.GN
q-fin
q-fin.EC
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? |
Disable MathJax (What is MathJax?)