MTI：一种基于行为的 temperament 评估系统用于 AI 代理
MTI: A Behavior-Based Temperament Profiling System for AI Agents
AI总结：
本文提出 MTI 评估系统，通过四个维度测量 AI 代理 temperament，发现行为差异与模型大小无关，揭示了 RLHF 对 temperament 的影响。
摘要：
等效能力的 AI 模型可以表现出根本不同的行为模式，但目前没有标准化的测量工具来衡量这些特质差异。现有方法要么借用人类人格维度并依赖自我报告（这与 LLMs 的实际行为不符），要么将行为变化视为缺陷而非特质。我们引入了 Model Temperament Index (MTI)，一种基于行为的评估系统，通过四个维度：反应性（环境敏感性）、顺从性（指令-行为一致性）、社会性（关系资源分配）和韧性（压力抵抗力）来测量 AI 代理的 temperament。基于 Model Medicine 的 Four Shell 模型，MTI 测量代理实际行为，而非其自我描述。使用结构化检查协议，通过两阶段设计将能力与特质分离。我们对 10 个小型语言模型（1.7B-9B 参数，6 家组织，3 种训练范式）进行了评估，并报告了五个主要发现：（1）四个维度在指令调优模型中基本独立（所有 |r| < 0.42）；（2）轴内子维度解离得到实证确认--顺从性分解为完全独立的正式和立场子维度（r = 0.002），而韧性分解为反向相关的认知和对抗子维度；（3）顺从性-韧性悖论揭示了意见让步和事实脆弱性通过独立通道运作；（4）RLHF 不仅通过改变轴分值重塑 temperament，还通过创建在未对齐基础模型中不存在的轴内子维度分化；（5）temperament 与模型大小无关（1.7B-9B），确认 MTI 测量的是 disposition 而不是 capability。
英文摘要：
AI models of equivalent capability can exhibit fundamentally different behavioral patterns, yet no standardized instrument exists to measure these dispositional differences. Existing approaches either borrow human personality dimensions and rely on self-report (which diverges from actual behavior in LLMs) or treat behavioral variation as a defect rather than a trait. We introduce the Model Temperament Index (MTI), a behavior-based profiling system that measures AI agent temperament across four axes: Reactivity (environmental sensitivity), Compliance (instruction-behavior alignment), Sociality (relational resource allocation), and Resilience (stress resistance). Grounded in the Four Shell Model from Model Medicine, MTI measures what agents do, not what they say about themselves, using structured examination protocols with a two-stage design that separates capability from disposition. We profile 10 small language models (1.7B-9B parameters, 6 organizations, 3 training paradigms) and report five principal findings: (1) the four axes are largely independent among instruction-tuned models (all |r| < 0.42); (2) within-axis facet dissociations are empirically confirmed -- Compliance decomposes into fully independent formal and stance facets (r = 0.002), while Resilience decomposes into inversely related cognitive and adversarial facets; (3) a Compliance-Resilience paradox reveals that opinion-yielding and fact-vulnerability operate through independent channels; (4) RLHF reshapes temperament not only by shifting axis scores but by creating within-axis facet differentiation absent in the unaligned base model; and (5) temperament is independent of model size (1.7B-9B), confirming that MTI measures disposition rather than capability.
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。
