arXiv cs.AI· Zhivar Sourati, Marianne Menglin Liu, Zheng Wang, Rongguang Wang, Sai Ashish Somayajula, Tao Sheng, Kyu J. Han, Sujith Ravi, Dan Roth·· 3 小时前AI 评分55
研究分析语义层、多智能体编排与记忆压缩对数据智能体的行为影响
Same Context, Different Agents: A Behavioral Study of Semantic Layers, Multi-Agent Orchestration and Memory Compaction in Data Agents
AI 导读
arXiv 论文（arXiv:2610.20959）系统分析语义层、编排和压缩三种上下文工程技术（CET），在五套基准的 1,514 个长程任务上测试七个 LLM。研究发现数据智能体分为调查者、过早回答者和提前放弃者三类行为组，CET 只有匹配行为组才能转化为准确率，对调查者反而可能有害，如编排使其过度自信损失 5 个百分点，组合使用 CET 的收益和损失均小于单独效果之和。
正文 · 原文
Abstract:Answering questions over heterogeneous data means combining evidence from files, databases, documents and the web. Data agents automate this, but their success depends on context. Three context-engineering techniques (CETs) help by shaping it: semantic layers describe the data, orchestration splits work among agents, and compaction condenses interaction history. Existing studies validate each CET alone, leaving open how agent behavior shapes its effect. We present a comprehensive analysis of the three CETs, alone and combined, with seven large language models (LLMs) on 1,514 long-horizon tasks from five benchmarks, covering behavior, accuracy and cost. We find that data agents without CETs form three behavior groups: investigators, premature answerers that answer without inspecting data, and early stoppers that give up. Each CET delivers its known benefit, from more investigation to smaller prompts, but only the behavior group it matches turns that benefit into accuracy, such as premature answerers under orchestration. For investigators, the same CETs can hurt: the semantic layer stops them from looking, orchestration makes them overconfident, costing 5 percentage points, and compaction makes them repeat steps. Combining CETs yields smaller gains and losses than their sum. We derive guidance for choosing CETs by agent behavior, task demands and accuracy-cost trade-offs.
| Comments: | v2: revised title |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.20959 [cs.CL] |
| (or arXiv:2610.20959v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20959 arXiv-issued DOI via DataCite |
Submission history
From: Zhivar Sourati [view email]
[v1]
Wed, 14 Oct 2026 21:29:59 UTC (1,195 KB)
[v2]
Thu, 22 Oct 2026 21:32:14 UTC (1,195 KB)
来源：arXiv cs.AI · arxiv.org