Mona Diab · page 4
Papers
- Processing Dialectal Arabic: Exploiting Variability and Similarity to Overcome Challenges and Discover Opportunities (2016 · 2 citations)
This is a one-page abstract of an invited talk at VarDial3, not a full research paper. Diab addresses the difficulty of processing informal Arabic, whose volume in text and speech has grown sharply, especially on social media. Such content matters for information extraction, political and commercial analytics, and… - Mechanistic Interpretability Should Prioritize Feature Consistency in Sparse Autoencoders (2026 · 1 citation)
Sparse autoencoders (SAEs) are a popular way to break LLM activations into sparse, human-readable features. - Personal Information Parroting in Language Models (2026 · 1 citation)
Web-scale pretraining corpora contain huge amounts of personal information (PI), and language models can reproduce it. - Arabic Dialect Processing Tutorial (2012 · 1 citation)
This is a tutorial abstract, not a research paper, so it reports no experiments, new data or measured results. - Figurative and Cultural Knowledge in LLMs: Investigating Cross-Domain Transfer through Fine-Tuning (2026)
Linguists and cognitive scientists have long argued that understanding figurative language depends on cultural familiarity. - Taming Object Hallucinations with Verified Atomic Confidence Estimation (2026)
Multimodal LLMs often describe objects, properties or relations that are not in the image. They are also easily swayed by small changes in how a question is worded. The authors propose TACO, a training-free pipeline that has the model check its own answers, with no external detectors or captioners. TACO works in four… - Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic X Communities (2026)
Most computational work on mental health in social media studies English speakers. This paper turns to Arabic and describes how people talk about three conditions in condition-specific X Communities: BPD, Bipolar and ADHD. The authors avoid diagnosing anyone. They frame the work as descriptive computational social… - Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic-Language X Communities (2026)
Most computational mental-health research examines English speakers and treats the task as detecting at-risk people. - Automatic Verification and Augmentation of Multilingual Lexicons (2016)
Large multilingual lexicons are usually built partly by hand and partly by automatic means, and checking their quality is normally done through costly manual review.
This page in Markdown → https://md.research.elamur.ai/free/en/authors/mona-diab/page/4/