4 papers
RRFC: Recursive Refinement via Feedback Conditioning for Iterative Image-to-Image Generation
Kareem Hassani, Chaymaa Abbas, Hadi Al Mubasher +1
Conditional image-to-image generators are single-shot: they map input features to an output in one forward pass and treat it as final, with no opportunity to improve on it. Althoug…
THESIS-MoE: Trainable Hierarchical Extraction and SteerIng of Sycophancy in Mixture-of-Experts
Kareem Hassani, Chaymaa Abbas, Lama Mawlawi +1
Sycophancy, the tendency of a language model to change its answer to match a user's stated belief, is a common alignment failure. Existing activation steering methods typically app…
Obscuring Data Contamination Through Translation: Evidence from Arabic Corpora
Chaymaa Abbas, Nour Shamaa, Mariette Awad
Data contamination undermines the validity of Large Language Model evaluation by enabling models to rely on memorized benchmark content rather than true generalization. While prior…
Can Small-Scale Data Poisoning Exacerbate Dialect-Linked Biases in Large Language Models?
Chaymaa Abbas, Mariette Awad, Razane Tajeddine
Style-conditioned data poisoning is identified as a covert vector for amplifying sociolinguistic bias in large language models. Using small poisoned budgets that pair dialectal pro…