Bolei Ma

Bolei Ma

@boleima.bsky.social

@LMU Munich

14 followers15 following7 posts

Bolei Ma
@boleima.bsky.social

๐Ÿ† ๐—•๐—ฒ๐˜€๐˜ ๐—ฃ๐—ฎ๐—ฝ๐—ฒ๐—ฟ ๐—”๐˜„๐—ฎ๐—ฟ๐—ฑ @ ๐—”๐—–๐—Ÿ ๐Ÿฎ๐Ÿฌ๐Ÿฎ๐Ÿฒ Thrilled to share that our linguistic paper "The Imperfective Paradox in LLMs" received a Best Paper Award at #ACL2026 ! ๐ŸŽ‰ Hoping this encourages more linguistics-informed evaluation in NLP. aclanthology.org/2026.acl-lon...

Bolei Ma
@boleima.bsky.social

Tomorrow, Wednesday, July 30, from 11:00-12:30 at #ACL2025 I am presenting our poster on surveying resources for understanding pragmatics in NLP models. Feel free to drop by and discuss! ๐Ÿ“Hall 4X, Poster board 374 aclanthology.org/2025.acl-lon...

Bolei Ma@boleima.bsky.social

๐Ÿค” Can LLMs read between the lines? Our another #ACL2025 paper surveys resources on how LLMs handle pragmatics like implicatures, deixis, and more. We map out a new landscape for both LLMs and linguistics in pragmatic research. ๐Ÿ“„ arxiv.org/abs/2502.12378 ๐Ÿง ๐Ÿ’ฌ #LLMs #Pragmatics

Bolei Ma
@boleima.bsky.social

I am presenting this paper today at 2PM in Room 1.62 (Session 3: IP-Orals). Feel free to attend and reach out if youโ€™d like to discuss ~ #ACL2025

Bolei Ma@boleima.bsky.social

๐ŸŽ‰ Our paper โ€œAlgorithmic Fidelity of LLMs in Generating Synthetic German Public Opinionsโ€ is accepted at #ACL2025 main conference as an oral presentation! ๐Ÿ‡ฉ๐Ÿ‡ช๐Ÿค– We study how well LLMs simulate real survey responses using open-ended German data, showing the left-leaning bias.

Bolei Ma
@boleima.bsky.social

๐Ÿค” Can LLMs read between the lines? Our another #ACL2025 paper surveys resources on how LLMs handle pragmatics like implicatures, deixis, and more. We map out a new landscape for both LLMs and linguistics in pragmatic research. ๐Ÿ“„ arxiv.org/abs/2502.12378 ๐Ÿง ๐Ÿ’ฌ #LLMs #Pragmatics

Bolei Ma
@boleima.bsky.social

๐ŸŽ‰ Our paper โ€œAlgorithmic Fidelity of LLMs in Generating Synthetic German Public Opinionsโ€ is accepted at #ACL2025 main conference as an oral presentation! ๐Ÿ‡ฉ๐Ÿ‡ช๐Ÿค– We study how well LLMs simulate real survey responses using open-ended German data, showing the left-leaning bias.

Bolei Ma
@boleima.bsky.social

๐Ÿ“„ arxiv.org/abs/2412.13169 ๐Ÿ—“๏ธ Monday, July 28, 14:00โ€“15:30 ๐Ÿ“Room 1.62, Session 3: IP-Orals Looking forward to seeing you in Vienna #ACL2025! ๐Ÿ‘‹ #LLMs #PublicOpinion #SurveyResearch w/ Berk @carohaensch.bsky.social @xinpeng.bsky.social Markus @barbaraplank.bsky.social Frauke @assenmacher.bsky.social

arxiv.orgAlgorithmic Fidelity of Large Language Models in Generating Synthetic German Public Opinions: A Case StudyIn recent research, large language models (LLMs) have been increasingly used to investigate public opinions. This study investigates the algorithmic fidelity of LLMs, i.e., the ability to replicate th...

You are up to date.