Skip to content
AI News HubLIVE
Original source2 min read

LLMs Silently Correct African American English: Auditing and Mitigating Dialect Bias via Activation Steering

Summary

Large language models (LLMs) systematically prefer Standard American English (SAE) continuations even when the preceding context is in African American English (AAE), effectively rewriting AAE into SAE. The authors propose an end-to-end framework to audit and mitigate this bias, including conditional Dialect Group Invariance (cDGI) and activation steering. They also release REAL-AAE, the largest real-AAE parallel corpus to date.

SourcearXiv Computational LinguisticsAuthor: Huan Wu, Ali Emami, Muhammad Furquan Hassan, Osaretin Igbinoba, Osakpolor Idusuyi, Osamede Igbinoba, Faiza Khan Khattak, Laleh Seyyed-Kalantari
LLMs Silently Correct African American English: Auditing and Mitigating Dialect Bias via Activation Steering
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 7 Jul 2026]

Title:LLMs Silently Correct African American English: Auditing and Mitigating Dialect Bias via Activation Steering

View a PDF of the paper titled LLMs Silently Correct African American English: Auditing and Mitigating Dialect Bias via Activation Steering, by Huan Wu and 7 other authors

View PDF HTML (experimental)

Abstract:African American English (AAE), a rule-governed dialect spoken by over 30 million people, is routinely misinterpreted and "corrected" by large language models (LLMs). Across six instruction-tuned LLMs (14B to 70B), we show that state-of-the-art models systematically prefer Standard American English (SAE) continuations even when the preceding context is in AAE, effectively rewriting AAE into SAE. We present an end-to-end framework to audit and mitigate this bias. For auditing, we introduce conditional Dialect Group Invariance (cDGI), which isolates true model bias from translator-induced artifacts, and a feature-level localization analysis that identifies which AAE markers most strongly trigger bias; we find that syntactic constructions, especially negative concord (e.g., "ain't nobody"), are universal triggers across all models. For mitigation, we introduce, to our knowledge, the first application of activation steering to dialect bias: a training-free, test-time method that extracts dialect directions via causal tracing and injects them into bias-relevant layers. Activation steering reduces bias 5 to 20 times more than prompting while preserving SAE fluency. To enable this work, we release REAL-AAE , the largest real-AAE parallel corpus to date: 17,479 AAE/SAE/ AAE_back triplets from natural tweets (2 to 6 times larger than prior real-AAE resources), validated automatically (BERTScore F1 = 0.95) and by three native AAE speakers (83.0% semantic agreement).

Subjects:

Computation and Language (cs.CL)

Cite as: arXiv:2607.06845 [cs.CL]

(or arXiv:2607.06845v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2607.06845

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Huan Wu [view email] [v1] Tue, 7 Jul 2026 22:36:30 UTC (6,512 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled LLMs Silently Correct African American English: Auditing and Mitigating Dialect Bias via Activation Steering, by Huan Wu and 7 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CL

new | recent | 2026-07

Change to browse by:

cs

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?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • LLMs exhibit a systematic preference for SAE over AAE, effectively correcting AAE input.
  • Conditional Dialect Group Invariance (cDGI) isolates true model bias from translator artifacts.
  • Activation steering reduces bias 5–20× more than prompting while maintaining SAE fluency.
  • REAL-AAE corpus contains 17,479 AAE/SAE/AAE_back triplets from natural tweets.

Highlights and analysis are generated automatically and may contain errors. Check the original source.

More recent stories on this topic · Models