智能体
arxiv.org ·
异源可解释性：探究大型语言模型的异质思维
该 arXiv 论文提出“异源可解释性（xeno-interpretability）”概念，主张 LLM 内部可能存在无法用现有人类概念（真实性、拒答、欺骗、人格、有害性等）充分表达的区分结构，称为 xeno-representations。作者区分人类可解释语义空间与模型原生语义空间，认为模型内部可能区分的空间远大于有限人类描述所能覆盖的范围，并提出可将实验识别与语义解释分离：即使语义内容无法用人类术语表达，内部表示仍可被可复现地定位、几何刻画、因果操控并关联到下游行为。论文据此勾勒识别 xeno-representations 的实证研究计划，并讨论其对 AI 安全与多智能体系统的含义——模型原生表示可能在交互智能体间传播和稳定，而通过人类可读通信仅部分可见。
AI推荐理由 arXiv 预印本，作者提出“异源可解释性（xeno-interpretability）”概念，区分人类可解释语义空间与模型原生表示空间，并给出识别异源表示（xeno-representations）的实验框架，讨论其对 AI 安全与多智能体系统的影响。属于概念性/立场性论文，摘要未提供具体实验数据或验证结果，价值在于提出新的可解释性研究议程，而非已证实的发现。
72
正文
Computer Science > Computation and Language
arXiv:2609.20408 (cs)
[Submitted on 17 Sep 2026]
Title:Xeno-Interpretability: Investigating the Alien Minds of LLMs
Authors:F. Pierucci, M. Bracale Syrnikov, M. Prandi, M. Galisai, F. Giarrusso, P. Bisconti
View a PDF of the paper titled Xeno-Interpretability: Investigating the Alien Minds of LLMs, by F. Pierucci and 5 other authors
View PDF
HTML (experimental)
Abstract:Large language models are usually interpreted through concepts that humans already possess: truthfulness, refusal, deception, personality, harmfulness, and related categories. This paper asks whether models may also represent and use distinctions for which no adequate human concept exists. We call such internal structures xeno-representations, and their study xeno-interpretability. We distinguish the human-interpretable semantic space from the xeno-semantic space: the region of model-native representations for which no adequate human conceptual counterpart is available. We show that the space of possible internal distinctions in an LLM is substantially larger than the space available through finite human descriptions. We then separate experimental identification from semantic interpretation: an internal representation may be reproducibly located, geometrically characterized, causally manipulated, and linked to downstream behaviour even when its semantic content cannot be adequately expressed in human terms. On this basis, we sketch an empirical programme to identify xeno-representations. We finally examine the implications for AI safety and multi-agent systems, where model-native representations may propagate and stabilize across interacting agents while remaining only partially visible through human-readable communication. Xeno-interpretability therefore shifts the aim of interpretability from finding human concepts inside models toward discovering and characterizing the representational structures that are native to the models themselves and might affect their behaviour in unpredictable ways.
Subjects:
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as:
arXiv:2609.20408 [cs.CL]
(or
arXiv:2609.20408v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2609.20408
Focus to learn more
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Matteo Prandi [view email]
[v1]
Thu, 17 Sep 2026 13:54:19 UTC (5,815 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Xeno-Interpretability: Investigating the Alien Minds of LLMs, by F. Pierucci and 5 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CL
new
|
recent
| 2026-09
Change to browse by:
cs
cs.AI
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
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, Media
Code, 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?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: 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?)