# HarmProfile：刻画前沿大语言模型的有害输出分布

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-06-11 08:00
- AIHOT 分数：74
- AIHOT 标记：精选
- AIHOT 链接：https://aihot.news/items/cmszbz3mk0969rodp45ep34o3
- 原文链接：https://arxiv.org/abs/2608.14577

## 精选理由

将有害输出视为可分析的对象而非单一攻击结果，80k 样本的风险画像为横向比较不同模型的安全边界提供了新基准。

## AI 摘要

HarmProfile 是一个以内容为中心的基准数据集，收集了 13 个模型家族、23 个前沿 LLM 的超过 80,000 个经验证的违规样本，覆盖 15 个危害类别和 57 个子类别。研究发现，前沿 LLM 能大规模稳定产生有害内容，且不同模型呈现不同的风险画像；危害性与多样性随模型能力增长，暗示模型可能在对齐表面之下隐藏着日益危险的知识。源代码已公开。

## 正文

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 .