前沿大语言模型（LLM）的安全评估在很大程度上将有害生成视为一种攻击结果，而非分析对象。因此，人们对模型在行为异常期间产生的有害输出知之甚少，部分原因在于大规模、高质量的前沿LLM异常行为数据集难以获取。为弥补这一空白，我们引入了HarmProfile，这是一个以内容为中心的基准数据集，收集了跨多种危害类别和模型家族的模型异常行为，并将由此产生的有害输出分布定义为模型层面的风险画像。其前提是，正如语言行为可以通过话语语料库来刻画一样，模型风险也可以通过其安全失败的内容、严重程度和变化来表征。HarmProfile包含来自13个模型家族、23个前沿LLM的超过80,000个经过验证的样本，涵盖15个危害类别和57个子类别。利用该语料库，我们发现前沿LLM确实能够大规模地产生有害内容，但表现出截然不同的风险画像；有害性和多样性均随模型能力的增强而增长，这表明前沿LLM可能表面上看起来安全，但在对齐表面之下却潜藏着日益危险的知识。我们的源代码可在 https://github.com/fresh-ma/HarmProfile 获取。
Frontier large language models (LLMs) safety evaluation has largely treated harmful generation as an attack outcome rather than as an object of analysis. Consequently, little is known about the harmful outputs produced during model misbehavior, partly because large-scale, high-quality collections of frontier-LLM misbehavior are difficult to obtain. To address this gap, we introduce HarmProfile, a content-centric benchmark dataset that collects model misbehavior across diverse harm categories and model families, and defines the resulting harmful-output distribution as a model-level risk profile. The premise is that, just as linguistic behavior can be characterized from an utterance corpus, model risk can be characterized from the content, severity, and variation of its safety failures. HarmProfile contains over 80,000 validated artifacts from 23 frontier LLMs across 13 model families, organized into 15 harm categories and 57 subcategories. Using this corpus, we find that frontier LLMs reliably produce harmful content at scale, yet exhibit distinct risk profiles; both harmfulness and diversity grow with model capability, suggesting that frontier LLMs may appear safe yet harbor increasingly dangerous knowledge beneath the alignment surface. Our source code is available at https://github.com/fresh-ma/HarmProfile .