Jul 2026
CHiPS: Character Histograms and Positional Signals for Lightweight Authorship Attribution in Romanian Texts
CHiPS, a lightweight character-level authorship attribution method for Romanian texts, requires no tokenization, syntactic analysis, pretrained language model, or transformer fine-tuning, and it avoids character $n$-gram features with $n \geq 2$ in the histogram component.
S. Avram, G. C. Ţurcaş
· arXiv.org · 0 citations