Dr. Chen Shani

My research interests are at the intersection of natural language processing (NLP) and human cognition. I aspire to improve language models by drawing inspiration from human cognition. In the past, I  focused on subjective human qualities such as humor, playfulness, personification, and creativity. Nowadays, I focus on shifting language models from the token- to the concept level. Meaning, token-based objectives are prone to spurious correlations. I address this problem by optimizing concept-aware language models that process content while ignoring superficial noise (see Research Agenda).

I am a post-doctoral Researcher at Stanford University, working with Professor Dan Jurafsky and Professor Jennifer Eberhardt.  

I will be joining Tel-Aviv University as an Assistant Professor (Senior Lecturer) in 2026.

I completed my Ph.D. under the supervision of Professor Dafna Shahaf at the Hebrew University of Jerusalem.  My Ph.D. focused on designing NLP architectures that incorporate principles from human cognition, with a specific interest in NLP tasks related to subjective human qualities: humor, playfulness, personification, and creativity. My findings demonstrate the potential of integrating human-inspired perspectives to advance NLP and create more human-like models.

I am also an ex-Amazonian, where I co-founded a team aimed to automatically enhance Alexa's sense of humor. In addition, I occasionally consult start-up companies.

Collaborations, thoughts, and brainstorming are my treat, so feel free to reach out!

My updated CV can be found here.