智能体
arxiv.org ·
@Maxm Pan
超越结果：面向高效智能体基准评测的双视角关系学习
智能体基准的评测成本显著高于常规 LLM 基准，基准压缩是自然解法，但现有方法主要建模任务—模型最终得分分布中的冗余。本文分析大规模轨迹，识别出六个与最终智能体表现系统性相关的互补过程信号，并提出 DualViewEval——一种联合利用结果关系与过程关系的智能体基准压缩方法，用于学习精确规模的最小集并预测全基准得分。在五个智能体基准和五个代表性基线上，DualViewEval 在所有数据集上取得最佳结果：仅用 20 个任务即在 APEX-Agents 和 BFCL 上实现 24–40 倍压缩，平均绝对误差较最强基线降低 14.5%–28.2%。
AI推荐理由 针对智能体基准评测成本远高于常规 LLM 基准的问题，指出既有压缩方法只建模任务—模型最终得分分布的冗余，进而分析大规模轨迹并识别出六个与最终智能体表现系统性相关的过程信号，提出联合利用结果与过程关系的 DualViewEval 压缩方法。在五个智能体基准和五个基线上均取得最佳结果，仅用 20 个任务即在 APEX-Agents 和 BFCL 上实现 24–40 倍压缩，MAE 降低 14.5%–28.2%。方法具体、指标明确，对智能体评测实践有直接价值；局限是摘要未说明过程信号的具体定义与跨基准泛化细节。
74
正文
Computer Science > Computation and Language
arXiv:2609.18909v1 (cs)
[Submitted on 16 Sep 2026]
Title:Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
Authors:Xinshuai Guo, Junjie Wu, Dolly Deng, Yinghui Li, Hai-Tao Zheng, Suncong Zheng, Maxm Pan
View a PDF of the paper titled Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking, by Xinshuai Guo and 6 other authors
View PDF
HTML (experimental)
Abstract:Agent benchmarks are substantially more costly to evaluate than conventional LLM benchmarks. Benchmark compression is therefore a natural solution, yet existing methods primarily model redundancy in task--model final-score distributions, which is important in agentic evaluation. To address this limitation, we analyze large-scale trajectories and identify six complementary process signals that are systematically associated with final agent performance. To disentangle agent performance redundancy from a complete perspective, we propose DualViewEval, an agent benchmark compression method that jointly exploits outcome and process relations to learn an exact-size miniset and predict the full-benchmark scores. Across five agent benchmarks and five representative baselines, DualViewEval achieves the best results in all datasets. With only 20 tasks, it achieves $24\times$--$40\times$ compression on APEX-Agents and BFCL, reducing mean absolute error (MAE) by $14.5\%$--$28.2\%$ over the strongest competitors while improving Kendall's $\tau$ by up to $7.2\%$ relative to EssenceBench on SWE-bench Verified. The selected minisets further reveal capability differences among different agents, providing compact and diagnostic feedback for efficient agentic model development.
Subjects:
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as:
arXiv:2609.18909 [cs.CL]
(or
arXiv:2609.18909v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2609.18909
Focus to learn more
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Xinshuai Guo [view email]
[v1]
Wed, 16 Sep 2026 16:49:50 UTC (1,150 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking, by Xinshuai Guo and 6 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CL
new
|
recent
| 2026-09
Change to browse by:
cs
cs.AI
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?)