arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.33775cs.AIcs.CL
收藏

System 1模型使用指南:超越大语言模型的快速决策

The Hitchhiker's Guide to System 1 Models: Fast Decision-Making Beyond LLMS

Technion – Israel Institute of Technology(以色列理工学院)

Haggai Roitman

AI总结:

本指南提出System 1决策模型的五家族分类法,分析其校准理论与现象,阐述其在智能体工作流中的集成方式,为相关领域研究者和工程师提供统一入口。

摘要:

大语言模型在推理和生成方面表现出色,但其自回归、逐token的推理方式导致速度过慢、成本过高,且对现代AI系统所需的许多常规决策(路由、门控、评分、分类)校准效果不佳,无法满足亚秒级延迟要求。一类新兴模型填补了这一空白:System 1决策模型,这类架构可在单次前向传播中将输入直接映射为类型化、结构化的决策。本指南提供了面向从业者的全面领域论述,提出了涵盖基于编码器的分类器、奖励模型、非自回归决策引擎、对比决策模型和提取模型的五家族分类法,并深入探讨了每个家族的架构、训练方法和部署权衡。核心主题是校准:我们推导了恰当评分规则的理论,分析了用于校准决策的强化学习(RLCD),并研究了近期发现的现象,包括IIA违反、序数尺度偏差和分布偏移下的少样本校准。在集成方面,我们描述了System 1模型如何作为智能体工作流中的决策基础设施,以及System 1与System 2的级联结构,该结构基于校准后的置信度在快速模型和慢速模型之间进行路由。本指南覆盖18个应用领域,梳理了快速发展的开源生态系统(数周内超过70个模型复现),全程提供可运行代码,包含从业者检查表、开源模型库以及22道带详细答案的自我评估题。本指南引用190篇文献,旨在作为快速决策、校准和生产AI系统交叉领域的研究人员与工程师的统一入口。

英文摘要:

Large Language Models excel at reasoning and generation but their autoregressive, token-by-token inference makes them unnecessarily slow, expensive, and poorly calibrated for the many routine decisions (routing, gating, scoring, classification) that modern AI systems require at sub-second latency. A growing class of models addresses this gap: System 1 decision models, architectures that map an input directly to a typed, structured decision in a single forward pass. This guide provides a comprehensive, practitioner-oriented treatment of the field. We propose a five-family taxonomy covering encoder-based classifiers, reward models, non-autoregressive decision engines, contrastive decision models, and extraction models, and we provide in-depth coverage of each family's architecture, training methodology, and deployment trade-offs. A central theme is calibration: we develop the theory of proper scoring rules, analyze Reinforcement Learning for Calibrated Decisions (RLCD), and examine recently discovered phenomena including IIA violations, ordinal-scale bias, and few-shot recalibration under distribution shift. On the integration side, we describe how System 1 models serve as decision infrastructure within agentic workflows and System 1 + System 2 cascades that route between fast and slow models based on calibrated confidence. The guide covers 18 application domains, maps the rapidly growing open-source ecosystem (over 70 model reproductions in weeks), provides runnable code throughout, and includes a practitioner's checklist, an open-source model zoo, and 22 self-assessment questions with detailed answers. Spanning 190 references, it is intended as a unified entry point for researchers and engineers working at the intersection of fast decision-making, calibration, and production AI systems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑