Topics: Food Science, Natural Language Processing, Data Mining, Knowledge Base Construction, Numeracy, Common Sense Reasoning

I am passionate about developing practical AI applications that encompass the culinary, biomedical, and clinical domains. Recently, large language models.

My current project is CoDiet, under the supervision of Dr. Joram Posma.

I have experience working with textual knowledge, which involves gathering and processing domain-specific data such as culinary recipes, general web content, review data, and structured information like blockchain transactions.I have employed traditional computational techniques, including search engines and relational and non-relational databases, to manage and store this diverse range of large-scale data.Throughout my Ph.D. tenure, my research has been centered around creating AI solutions that capture semantic meaning. This involves various aspects, including enhancing language models to better understand numeracy, implementing AI systems that generalize statistical methods for culinary recommendations, and developing reinforcement learning systems within the finance domain for portfolio selection.My academic projects have been closely aligned with industry. I have completed an internship and collaborated on projects with Sony AI, in addition to founding startups.

[Preprint]

<Small Language Models Learn Enhanced Reasoning Skills from Medical Textbooks>
Hyunjae Kim, Hyeon Hwang, Jiwoo Lee, Sihyeon Park, Dain Kim, Taewhoo Lee, Chanwoong Yoon, Jiwoong Sohn, Donghee Choi, Jaewoo Kang
* Paper: https://arxiv.org/abs/2404.00376
* Huggingface: https://huggingface.co/dmis-lab/meerkat-7b-v1.0