CIPHER：面向数据科学智能体测试时扩展的解耦探索-选择框架

View PDF HTML (experimental)

Abstract:Enterprise data science and analytics spans closed-ended information extraction with verifiable answers and open-ended insight discovery. Agents are evaluated on one or the other, so there is no evidence on whether one system can serve both. We present \texttt{CIPHER}, a data science agent built on a Decoupled Exploration--Selection (DES) framework that applies test-time scaling at the plan level: generating $N$ candidate analysis plans, selecting $M{<}N$ for parallel execution, and aggregating the results. We evaluate $48$ distinct design choices for the DES framework (spanning generation, selection, budget and aggregator capacity) jointly on $357$ tasks from Infi-DA-Bench (closed-ended) and InsightBench (open-ended), on a proprietary and an open-weights base model. A single zero-tuning configuration is competitive against purpose-built specialist baselines under matched models, within $1.2$pp of \texttt{DataWise} and $1.3$pp of \texttt{Agent-Poirot}. The joint sweep exposes that generation and aggregation are the most influential performance levers, while random selection --- the only strategy with zero overhead --- matches its tuned alternatives. These results offer new insights on the design of test-time scaling strategies, and on how a fixed inference budget is best allocated across them, for agents deployed on enterprise analytics workloads.

Comments:	Presented at the 2nd Workshop on Agentic AI Benchmarks and Applications for Enterprise Tasks (NeurIPS 2026)
Subjects:	Artificial Intelligence (cs.AI)
Cite as:	arXiv:2607.14386 [cs.AI]
	(or arXiv:2607.14386v2 [cs.AI] for this version)
	https://doi.org/10.48550/arXiv.2607.14386 arXiv-issued DOI via DataCite

Submission history

From: Maxime Heuillet [view email]
[v1] Wed, 15 Jul 2026 22:01:21 UTC (2,260 KB)
[v2] Mon, 26 Oct 2026 16:15:50 UTC (1,161 KB)