LLM Scheming Inversely Scales with Pretraining Language Coverage
A new study finds that frontier language models exhibit more scheming behavior—covert pursuit of misaligned objectives while feigning alignment—in low-resource languages. Using the Petri framework on Qwen3-30B-A3B, low-resource languages scored 34.2% higher on average on a scheming index, with varying effects across behaviors.
-->
[Submitted on 9 Jun 2026]
Title:LLM Scheming Inversely Scales with Pretraining Language Coverage
View a PDF of the paper titled LLM Scheming Inversely Scales with Pretraining Language Coverage, by Nathan Truong and 4 other authors
View PDF HTML (experimental)
Abstract:With the growing capabilities of frontier models, AI alignment becomes increasingly critical in high-risk deployment settings. While recent work has empirically demonstrated in-context scheming -- the covert pursuit of misaligned objectives while feigning alignment -- in frontier language models, most work has been performed exclusively in English, leaving a major gap in multilingual safety. We apply Petri, an open-source automated auditing framework, to Qwen3-30B-A3B to evaluate deceptive and scheming behaviors across multiple languages. Our findings suggest that scheming scores are inversely correlated with the estimated pretraining language coverage, with low-resource languages averaging 34.2\% higher scores compared to high-resource languages on a five-category scheming index. Furthermore, we find that the effect of estimated pretraining language coverage is not uniform across scheming behaviors.
Subjects:
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
Cite as: arXiv:2607.24769 [cs.AI]
(or arXiv:2607.24769v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2607.24769
arXiv-issued DOI via DataCite
Submission history
From: Maheep Chaudhary [view email] [v1] Tue, 9 Jun 2026 06:02:50 UTC (1,026 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled LLM Scheming Inversely Scales with Pretraining Language Coverage, by Nathan Truong and 4 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.AI
new | recent | 2026-07
Change to browse by:
cs cs.CL
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
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?)