Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs

• URL: http://arxiv.org/abs/2607.03936v1
• Date: Sat, 04 Jul 2026 16:04:44 GMT
• Title: Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs
• Abstract summary: Key challenge in Arabic NLP is the scarcity of dialectal data relative to Modern Standard Arabic (MSA)<n>From an interpretability perspective, this raises a fundamental question: where and how are dialectal features encoded within model internals?<n>This study investigates two complementary inference-time approaches that serve simultaneously as interpretability probes and control mechanisms.
• Score: 60.9510933247844
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: A key challenge in Arabic NLP is the scarcity of dialectal data relative to Modern Standard Arabic (MSA), causing LLMs to overproduce MSA and struggle with dialectally accurate generation. From an interpretability perspective, this raises a fundamental question: where and how are dialectal features encoded within model internals, and can these representations be leveraged to improve dialect generation without fine-tuning? This study investigates two complementary inference-time approaches that serve simultaneously as interpretability probes and control mechanisms. First, we conduct a neuron-level analysis, identifying sparse neuron populations that encode dialect-specific features and showing that amplifying or suppressing these neurons can steer model outputs toward target dialects. Second, motivated by the entanglement of dialectal features at the single-neuron level, we apply a vector-steering approach that extracts dialect-specific activation directions and injects them during inference. Together, these methods illuminate the geometry of dialectal knowledge in Arabic LLMs and offer a principled, interpretability-grounded framework for dialect control without requiring dialect-specific fine-tuning.

Related papers

• Can LLM Agents Identify Spoken Dialects like a Linguist? [9.242671426685725]
  We explore the ability of large language models (LLMs) as agents in understanding the dialects and whether they can show comparable performance to models such as HuBERT in dialect classification.<n>Our approach uses phonetic transcriptions produced by ASR systems and combines them with linguistic resources such as dialect feature maps, vowel history, and rules.
  arXiv  Detail & Related papers  (2026-03-31T10:24:20Z)
• How does Alignment Enhance LLMs' Multilingual Capabilities? A Language Neurons Perspective [64.79894853375478]
  We propose a new finer-grained neuron identification algorithm, which detects language neurons(including language-specific neurons and language-related neurons) and language-agnostic neurons.<n>Based on the distributional characteristics of different types of neurons, we divide the LLMs' internal process for multilingual inference into four parts.<n>We systematically analyze the models before and after alignment with a focus on different types of neurons.
  arXiv  Detail & Related papers  (2025-05-27T17:59:52Z)
• Mechanistic Understanding and Mitigation of Language Confusion in English-Centric Large Language Models [56.61984030508691]
  We present the first mechanistic interpretability study of language confusion.<n>We show that confusion points (CPs) are central to this phenomenon.<n>We show that editing a small set of critical neurons, identified via comparative analysis with a multilingual-tuned counterpart, substantially mitigates confusion.
  arXiv  Detail & Related papers  (2025-05-22T11:29:17Z)
• DialUp! Modeling the Language Continuum by Adapting Models to Dialects and Dialects to Models [26.91517488888121]
  Most of the world's languages and dialects are low-resource, and lack support in machine translation (MT) models.<n>We present DialUp, consisting of a training-time technique for adapting a pretrained model to dialectal data.<n>We show considerable performance gains for several dialects from four language families, and modest gains for two other language families.
  arXiv  Detail & Related papers  (2025-01-27T23:53:04Z)
• What Do Dialect Speakers Want? A Survey of Attitudes Towards Language Technology for German Dialects [60.8361859783634]
  We survey speakers of dialects and regional languages related to German. We find that respondents are especially in favour of potential NLP tools that work with dialectal input.
  arXiv  Detail & Related papers  (2024-02-19T09:15:28Z)
• Natural Language Processing for Dialects of a Language: A Survey [56.93337350526933]
  State-of-the-art natural language processing (NLP) models are trained on massive training corpora, and report a superlative performance on evaluation datasets.<n>This survey delves into an important attribute of these datasets: the dialect of a language.<n>Motivated by the performance degradation of NLP models for dialectal datasets and its implications for the equity of language technologies, we survey past research in NLP for dialects in terms of datasets, and approaches.
  arXiv  Detail & Related papers  (2024-01-11T03:04:38Z)
• Content-Localization based System for Analyzing Sentiment and Hate Behaviors in Low-Resource Dialectal Arabic: English to Levantine and Gulf [5.2957928879391]
  This paper proposes to localize content of resources in high-resourced languages into under-resourced Arabic dialects. We utilize content-localization based neural machine translation to develop sentiment and hate classifiers for two low-resourced Arabic dialects: Levantine and Gulf. Our findings shed light on the importance of considering the unique nature of dialects within the same language and ignoring the dialectal aspect would lead to misleading analysis.
  arXiv  Detail & Related papers  (2023-11-27T15:37:33Z)
• Quantifying the Dialect Gap and its Correlates Across Languages [69.18461982439031]
  This work will lay the foundation for furthering the field of dialectal NLP by laying out evident disparities and identifying possible pathways for addressing them through mindful data collection.
  arXiv  Detail & Related papers  (2023-10-23T17:42:01Z)
• Post-hoc analysis of Arabic transformer models [20.741730718486032]
  We probe how linguistic information is encoded in the transformer models, trained on different Arabic dialects. We perform a layer and neuron analysis on the models using morphological tagging tasks for different dialects of Arabic and a dialectal identification task.
  arXiv  Detail & Related papers  (2022-10-18T16:53:51Z)
• Interpreting Arabic Transformer Models [18.98681439078424]
  We probe how linguistic information is encoded in Arabic pretrained models, trained on different varieties of Arabic language. We perform a layer and neuron analysis on the models using three intrinsic tasks: two morphological tagging tasks based on MSA (modern standard Arabic) and dialectal POS-tagging and a dialectal identification task.
  arXiv  Detail & Related papers  (2022-01-19T06:32:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.