超越表面形式：面向机制的间接语言编码综合分类法——用于基于LLM的编码语言检测
评分
53精选理由
论文提出了创新性且经过实证验证的ILE分类法，在LLM检测中取得显著性能提升，对内容安全领域有重要实用价值和方法论贡献。
AI 点评
「该分类法为LLM检测不断演变的网络暗语和规避性语言提供了系统化的认知框架，有望提升社交平台的内容审核准确性与适应性。」
速览
社交媒体用户常使用间接语言表达（ILE）规避内容审核，如算法暗语、委婉语和对抗性混淆。现有分类法侧重于沟通意图，导致检测不稳定。本文提出一种面向机制的综合分类法，抽象出编码底层操作，并将其集成到LLM提示中。在2000条手动标注的TikTok和Bluesky帖子上，新分类法在文档级和跨度级性能均最强，准确率提升4.7%，F1提升5.4%，为检测新兴编码语言提供了稳定支架。
- 提出面向机制的间接语言表达（ILE）综合分类法，专注于编码的底层操作而非沟通意图。
- 将该分类法集成到LLM提示中，在2000条手动标注的社交媒体帖子上评估。
- 与四种现有分类法和无分类法基线相比，准确率提升4.7%，F1提升5.4%。
- 揭示了全面、面向机制的分类法作为检测新兴编码语言的稳定支架的重要性。
正文
Computer Science > Computation and Language
arXiv:2606.27314v1 (cs) [Submitted on 25 Jun 2026]Title:Beyond Surface Forms: A Comprehensive, Mechanism-Oriented Taxonomy of Indirect Linguistic Encoding for LLM-Based Coded Language Detection
Authors:Hamid Reza Firoozfar, Mohammadsadegh Abolhasani, Reza Mousavi, Paul Jen-Hwa Hu View a PDF of the paper titled Beyond Surface Forms: A Comprehensive, Mechanism-Oriented Taxonomy of Indirect Linguistic Encoding for LLM-Based Coded Language Detection, by Hamid Reza Firoozfar and 3 other authors View PDFAbstract:To avoid moderation and surveillance on social media, some users routinely invent indirect linguistic expressions (ILE) that camouflage sensitive meanings. Such expressions surface as algospeak, euphemisms, and adversarial obfuscation, depending on intent and context, and they involve recurring encoding mechanisms. We propose a comprehensive, mechanism-oriented taxonomy of ILE that abstracts away from communicative goals and instead categorizes the underlying operations through which meaning is encoded and recovered. We evaluate the taxonomy by incorporating it into LLM prompts and comparing it with four existing taxonomies and a no-taxonomy baseline, using 2,000 manually annotated TikTok and Bluesky posts. The proposed taxonomy attains the strongest document- and span-level performance across the three LLMs, achieving an improvement of 4.7% in accuracy and 5.4% in F1 over the best-performing benchmark. The empirical results reveal the importance of a comprehensive, mechanism-oriented taxonomy as a stable scaffold for detecting emerging coded language and a useful input to content moderation. Disclaimer: This paper contains content that may be profane, vulgar, or offensive.
| Comments: | Submitted for review in ARR for EMNLP 2026 |
| Subjects: | Computation and Language (cs.CL) |
| ACM classes: | I.2.7 |
| Cite as: | arXiv:2606.27314 [cs.CL] |
| (or arXiv:2606.27314v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.27314 Focus to learn more arXiv-issued DOI via DataCite |
Submission history
From: Mohammadsadegh Abolhasani [view email][v1] Thu, 25 Jun 2026 17:29:45 UTC (2,055 KB)
Full-text links:
Access Paper:
-
View a PDF of the paper titled Beyond Surface Forms: A Comprehensive, Mechanism-Oriented Taxonomy of Indirect Linguistic Encoding for LLM-Based Coded Language Detection, by Hamid Reza Firoozfar and 3 other authors
- View PDF
- TeX Source
Current browse context:
cs.CL < prev | next >new | recent | 2026-06 Change to browse by: cs
References & Citations
export BibTeX citation Loading...BibTeX formatted citation
× Data provided by:Bookmark
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, MediaCode, Data and Media Associated with this Article
alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) DemosDemos
Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related PapersRecommenders and Search Tools
Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) About arXivLabsarXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)相关条目
85
精选Setoka：面向异构数据个性化Agent的层级用户理解基准
这篇论文非常值得关注，尤其对于从事个性化Agent、对话系统和记忆增强AI的研究者和工程师。它不仅仅是一个基准，而是从心理学出发定义了用户理解的层级，揭示了现有系统在高层理解上的系统性失败。如果你在设计基于记忆的Agent，这篇文章会提醒你记忆系统需要支持跨时间、跨来源的抽象推理，而不仅仅是检索。
85
精选BAND：高维分布估计的贝叶斯网络回归方法——打破维度诅咒
这篇论文值得所有从事高维统计、概率建模和生成式AI的研究者关注。它从理论上给出了一个简单却有力的答案：稀疏结构能打破维度诅咒，而贝叶斯网络恰好提供了自然框架。相比VAE/GAN这类黑箱方法，BAND可解释性强且收敛有保证，尤其适合工业界对可靠置信区间的需求。如果你是研究时间序列预测或风险管理的，这篇论文可能会改变你处理高维密度估计的方式。
84
精选循环一致与不确定性感知的托卡马克边缘等离子体神经代理模型
这篇论文值得关注，因为它展示了如何将先进的AI技术（循环一致性、不确定性量化）直接用于解决核聚变中的核心计算瓶颈。对于聚变领域的研究人员和工程师，这是一个能在不牺牲精度的前提下将模拟速度提升数个数量级的实用工具；对于AI从业者，它提供了一个科学应用中自监督学习和逆问题求解的优秀案例。
83
精选等变学习可迁移的三维经典密度泛函
这篇论文最值得关注的点在于，它把机器学习和统计力学的核心难题直接对接，用等变网络学出了一个可迁移的密度泛函，而且不依赖任何标签。对于做液体物理或软物质的同行，这可能是新的研究范式：以后是不是不用再针对每个体系从头做模拟了？对于AI研究者，这也展示了如何将物理对称性和变分原理嵌入模型，而不是单纯拟合数据。建议重点读一下方法部分，尤其是等变性和变分一致性的实现方式，那才是真正的精髓。
83
精选c-Rectified Flow 的计算与统计保证
这篇论文真正把整流流从“经验好用”推向了“理论可控”，对做生成模型底层算法的人尤其重要。如果你在调 FLUX 或 SD3 这类管线，理解普通整流流在协方差不可交换时可能失效，能帮你避开一些隐蔽的训练陷阱。
83
精选从高保真UMI数据学习可部署操作策略
这篇论文最亮眼的地方是证明了纯无机器人数据也能达到与真实遥操作相当的性能，彻底打破了后训练阶段必须依赖少量真实数据的惯例。对于想降低硬件成本的研究组和工业机器人应用开发者来说，HiFi-UMI的数据生产方案和开源数据集是直接可用的资源。建议关注其头戴式SLAM和微秒同步的工程细节，这些是精度达标的关键。
83
精选资本市场LLM可靠性评分：从“看似合理”到“可银行化”
这篇论文对金融圈和AI圈都很有价值：它给出了一个可操作的、面向实际工作流的评估框架，比泛泛的基准更贴近真实需求。如果你在金融科技公司做LLM落地，或者负责模型的“可解释性”和“可审计性”，这篇值得细读。决策有用性维度的发现尤其值得关注——评审们一致认为这是关键，但不同模型表现差异最大。
83
精选优化器状态应放在哪里？面向内存高效MoE训练的分层状态分配
这篇论文对训练MoE模型的团队非常有价值：它证明了通过精细的优化器状态分配，可以在不牺牲精度的情况下大幅降低显存占用，甚至还能提升性能。尤其适合那些受限于40GB或80GB GPU显存的实验室或企业。值得仔细阅读其分层策略和动量保留的消融分析，可为实际训练框架（如DeepSpeed、Megatron）提供优化方向。