Do LLMs differentiate linguistic traits internally?: A layer-wise analysis

Samhita Bollepally, Takashi Yamauchi, Manju Muralee +1

Large language models often produce judgments that resemble human responses, but it remains unclear whether they internally capture linguistic nuances such as conventional versus i…