Research

Publications & talks

Mostly on the interpretability of language models from a linguistics-oriented perspective. Newest first.

  1. 2026 Under review

    Lexical frequency and grammatical generalization in LLMs

    Under review (ARR 2026 / EMNLP 2026)

    Investigates how lexical frequency affects grammatical preferences in LLMs, specifically whether models stay robust when minimal pairs involve rare or uncommon lexical items.

  2. Feb 2026

    Large Language Models Are Robust to Low-Frequency Words in Grammatical Evaluation

    言語処理学会 (NLP) 2026, poster

    A smaller, earlier version of the grammatical-generalization work, presented as a poster.

Research interests

mechanistic interpretabilitylarge language modelslinguistic representations in neural networksgrammatical generalizationsyntaxlexical frequencymultilingualitysparse autoencodersrepresentation analysisNLP evaluation