The paper investigates how Arabic dialects are represented in large language models and whether they can be steered at inference time. By analyzing neuron-level sparsity and vector steering, the authors find that only a small fraction of neurons encode dialect-specific features, while distributed activation directions are more effective for steering. Vector steering can induce dialectal output from both dialectal and MSA prompts, whereas neuron steering works only when the prompt is already dialectal.
arXiv:2609.23955v1 Announce Type: new Abstract: Previous work on Egyptian Arabic in NLP has focused largely on the prestigious Cairene Egyptian Arabic (CEA) dialect, resulting in a lack of representa...
Dialectal variation remains a major challenge for multilingual language models. Perturbation-based continued pre-training (CPT) has emerged as a promising approach to improving robustness, yet existing work largely evaluates individual perturbation strategies in isolation and provides limited insight into why they work.
arXiv:2601. 22594v2 Announce Type: replace-cross Abstract: The high-level concepts that a neural network uses to perform computation need not be aligned to individual neurons (Smolensky, 1986).
arXiv:2609.07037v1 Announce Type: new Abstract: Activation steering has emerged as a lightweight, inference-time approach to control the behavior of Large Language Models (LLMs). However, traditional...
arXiv:2604. 03532v2 Announce Type: replace-cross Abstract: Large language models (LLMs) show strong multilingual capabilities, yet reliably controlling the language of their outputs remains difficult.
arXiv:2507. 18043v2 Announce Type: replace-cross Abstract: Inference-time steering methods offer a lightweight alternative to fine-tuning large language models (LLMs) and vision-language models (VLMs) by modifying internal activations at test time without updating model weights.
arXiv:2608. 12334v1 Announce Type: cross Abstract: Despite the impressive multilingual capabilities of Large Language Models, the latent dynamics dictating language selection remain poorly understood.
arXiv:2609.07876v1 Announce Type: cross Abstract: Recent methods in language model interpretability employ techniques such as sparse autoencoders to decompose residual stream contributions into linea...
arXiv:2607. 00510v1 Announce Type: new Abstract: Knowing which training examples drive outputs is fundamental to auditing, correcting, and understanding language models, yet for modern LLMs this remains expensive, approximate, and largely post-hoc.
arXiv:2609.14151v1 Announce Type: cross Abstract: Steering language models using a set of contrastive representations has been a canonical and computationally efficient method for controlling model b...
Sparse autoencoders (SAEs) are proposed to extract numerous features from large language model (LLM) representations, yet explaining these features still relies primarily on external observation. This reliance leads to superficial explanations inferred from observed model behavior and computational inefficiency from collecting such behavioral evidence at scale.
Large language models (LLMs) trained on next-token prediction exhibit remarkable in-context learning (ICL) abilities, yet the representations that support ICL remain poorly understood. We consider suc...