LAB LOGA page on LabLog, a review site (not the lab's official page)
Kawahara Lab
Machine-translated · View original
A lab that builds AI that can understand and use language the way humans do.
AboutAbout the lab
Building AI that understands language
Kawahara Lab studies artificial intelligence systems that can understand and use language as well as humans do.
The lab studies building, using, and evaluating large language models (LLMs), as well as methods for incorporating knowledge into LLMs.
It also puts effort into building evaluation datasets for Japanese language processing, such as Japanese language understanding benchmarks.
The lab draws on insights from both computer science and linguistics.
Key terms
- Large language model
- An artificial intelligence model that handles human language, trained on large amounts of text data
- Benchmark
- A shared collection of problems or datasets used to measure the performance of AI
- Word sense ambiguity
- The property that the same word can be interpreted as having different meanings depending on context
NewsLab news
A student received the Best Paper Award of the JSAI 40th Anniversary Paper Award from the Japanese Society for Artificial Intelligence (JSAI)
A student received the Young Researcher Encouragement Award and the Committee Special Award at the Annual Meeting of the Association for Natural Language Processing (NLP2026)
A master's student received an Excellence Award at the JSAI Annual Conference
A student received an Encouragement Award and other awards at the NLP Young Researchers' Association (YANS) Symposium
A student received the IPSJ Yamashita SIG Research Award from the Information Processing Society of Japan
ResearchResearch themes
The lab studies artificial intelligence systems that can understand and use language at a human level. It works on building, using, and evaluating large language models (LLMs), methods for incorporating knowledge into LLMs, and building benchmarks.
*Illustrative imageJapanese Language Understanding Benchmark
Language understanding benchmarks had centered on English, so the lab built a Japanese counterpart (JGLUE), making it possible to evaluate Japanese language models.
*Illustrative imageDetecting Hard-to-Read Sentences
Focusing on word sense ambiguity, where the same word takes on different meanings depending on context, the lab studies automatic detection of hard-to-read sentences. This research received an excellence award at an academic conference.
*Illustrative imageJapanese Truthfulness Benchmark
To check whether the outputs of large language models are correct, the lab built a Japanese truthfulness benchmark (JTruthfulQA) and uses it to evaluate models.
*Illustrative imageWording and AI Performance
The lab tests how polite and impolite wording affect the performance of large language models, in English, Chinese, and Japanese.
MembersFaculty
The members of Kawahara Lab consist of faculty, doctoral, master's, and undergraduate students, and administrative staff.As of 2026
- ProfessorDaisuke Kawahara
VoicesStudent voices
Ratings and reviews reflect the experiences of individual current or former students, not the lab's official views. Whether a lab has reviews, or how many, says nothing about the lab itself.
Students who take part in an interview may receive a thank-you gift worth ¥1,000, regardless of what they say.
Student ratings (0 students)
--
--
--
--
--
--
Student voices answer questions like these
- What's great about this lab?
- What did you learn after joining?
- What's the vibe, and what is the professor like?
- Who is this lab for?
- Where do graduates go?
Your experience can help younger students choose their lab. Your name won't be shown.
Write a lab reviewAnonymous · Your draft is saved if you stopRelatedLabs similar to Kawahara Lab
Kobayashi LabSchool of Fundamental Science and Engineering, Department of Communications and Computer EngineeringSpeech recognition, image recognition, natural language proc…
Nakazato LabSchool of Fundamental Science and Engineering, Department of Communications and Computer EngineeringNetwork information processing, future Internet
Ogawa LabSchool of Fundamental Science and Engineering, Department of Communications and Computer EngineeringSpoken language processing, acoustic signal processing, patt…
Katto LabSchool of Fundamental Science and Engineering, Department of Communications and Computer EngineeringImage processing; information networksThe information on this page is compiled by the LabLog team from the lab's public information and interviews. For corrections or removal requests, please contact LabLog, not the lab.
Report an errorAbout the images
The top and research theme images are illustrative and were not taken at the lab. They are cropped from Wikimedia Commons images. Cropped CC BY-SA images are provided under the same CC BY-SA license as the original (version shown on each line).
- Top image: “Kanji (Unsplash)” Niketh Vellanki / CC0 (cropped)
- Japanese Language Understanding Benchmark: “Prototype kilogram replica” Japs 88 / CC BY-SA 3.0 (cropped)
- Detecting Hard-to-Read Sentences: “Daqing Science and Technology Museum optical illusion pillars” Jason Zhang / CC BY-SA 3.0 (cropped)
- Japanese Truthfulness Benchmark: “Balance scale MET DP318014” Henry N. Hooper and Company / O.P. Drake / CC0 (cropped)
- Wording and AI Performance: “Tea bowl (chawan), Japan, Edo period, 18th-19th century, earthenware, black glaze (raku) - Montreal Museum of Fine Arts - Montreal, Canada - DSC09533” Daderot / Public domain (cropped)
- Related lab (Kobayashi Lab): “Spectrogram of Papuan Malay sentence de bicara keras” Emflazie / CC0 (cropped)
- Related lab (Nakazato Lab): “Network cables in server room” ProjectManhattan / CC BY-SA 3.0 (cropped)
- Related lab (Ogawa Lab): “Spectrogram - It Rains a Lot in Portland” Aaron Parecki / CC BY 2.0 (cropped)
- Related lab (Katto Lab): “Studio A broadcast area ( )” Atomic Taco from Seattle, WA, USA / CC BY-SA 2.0 (cropped)
About LabLog
LabLog is a lab review app for students only. Student members at each university interview students in labs and publish what they say. Student voices can only be read by students who sign up with a university email address. Operated by bestiee Inc.
Learn more about LabLog ↗Search engines and AI may reference the basic lab info and rating summaries on this page and the description of LabLog. Student voices are for student members only. Automated collection by programs (scraping, or input to or training of AI) is prohibited under Article 8 of the Terms of Service. If you need data for research or analysis, please contact us.