跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 9 天前AI 評分66

更多智能體、更多樣本,還是更多思考?拆解多智能體大語言模型中的協調、集成與測試時計算

More Agents, More Samples, or More Thinking? Untangling Coordination, Ensembling, and Test-Time Compute in Multi-Agent LLMs

AI 導讀

這項研究綜合九項主要實證評估、共 516 種多智能體配置,發現與樣本數或資源預算相匹配的基線相比,多智能體系統準確率平均低 0.042。在這些可比對照中,74% 的多智能體結果準確率等於或低於基線;由單次貪心基線改用抽樣基線,準確率差異平均下降 0.038。

正文

Published on Nov 13

Authors:

Abstract

Evaluations of multi-agent language models often compare communicating architectures against a single greedy pass. This practice confounds coordination with sample multiplicity and inference compute. We synthesize nine primary empirical evaluations comprising 516 unique multi-agent configurations organized along a resource vector R with six dimensions and a baseline ladder with four categories. Against baselines with a single pass (Category A), multi-agent accuracy differences are negative overall and near zero when omitting the largest contributing study. When evaluated against baselines matched on sample count (Category B) or resource budget (Category C), multi-agent systems average -0.042 in accuracy, and multilevel random effects modeling on the log odds scale remains negative across sensitivity checks. Multi-agent accuracy equals or falls below the baseline in 74% of these comparable contrasts, and study means are nonpositive in almost all evaluated studies. Across studies reporting both baselines, moving from a greedy pass to a sampled baseline shifts accuracy differences by -0.038 on average. Because communication adds no information when agents share identical evidence, positive returns require partitioned evidence, external verification, or error diversity across models. We provide a reviewer checklist and accounting utilities.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2611.17610

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2611.17610 in a model README.md to link it from this page.

Datasets citing this paper 2

sajjadi/pace-contrasts

Viewer •

Updated 1 day ago

•

1.36k

•

36

sajjadi/hidden-profile-control

Viewer •

Updated 1 day ago

•

500

•

25

Spaces citing this paper 1

Collections including this paper 1

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co