Look Who’s Talking: Stanford Linguists at EACL 2023

The 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2023), held in Dubrovnik from May 2 to 6, featured the work of many Stanford linguists:

• Dan Jurafsky co-authored a main paper titled “When Do Pre-Training Biases Propagate to Downstream Tasks? A Case Study in Text Summarization”.
• Erika Petersen and Christopher Potts authored a long findings paper titled “Lexical Semantics with Large Language Models: A Case Study of English break”.
• Computer Science Ph.D. student Tolúlọpẹ́ Ogunremi, Dan Jurafsky, and Chris Manning presented a long findings paper titled “Mini But Mighty: Efficient Multilingual Pretraining with Linguistically-Informed Data Selection”.

Stanford linguists also presented during the 5th Workshop on Research in Computational Linguistic Typology and Multilingual NLP, held by the ACL Special Interest Group on Typology (SIGTYP 2023) on May 6 as part of the EACL conference.

• Computer Science Ph.D. student Isabel Papadimitriou, Linguistics undergraduate Kezia Lopez, and Dan Jurafsky presented an EACL short findings paper titled “Multilingual BERT has an accent: Evaluating English influences on fluency in multilingual models”. 
• Alum Ciyang Qing (Ph.D. ’20) co-presented a paper titled “A Crosslinguistic Database for Combinatorial and Semantic Properties of Attitude Predicates”.