[Adapt] [Seminar] Unveiling A Core Linguistic Region in Large Language Models

[hajin]

Hi, Adapters,Over the years, the field of NLP has been at the forefront of understanding the core principles of intelligence. The emergence of LLMs showcases a significant breakthrough. However, the academic community lacks a systematic understanding of the internal mechanisms of LLMs’ intelligence. Nevertheless, insights from cognitive science may offer fresh perspectives on this matter. 


In this talk, I will introduce a recent paper titled "Unveiling A Core Linguistic Region in Large Language Models". Inspired by cognitive science research, this paper investigates whether specific functional regions exist within LLMs. They identify a core region specifically responsible for language processing within LLMs. This region occupies only about 1% of the model’s parameters but plays a crucial role in maintaining the overall linguistic competence of the model.




Hope you find this talk useful.




Time: Wed 10 am. - 11:30 am.
Meeting link: https://teams.microsoft.com/l/meetup-join/19%3ameeting_M2VmMTU5MzgtODUzOC00NmU4LTg0MzktNGFjNDdiMmIwYTI1%40thread.v2/0?context=%7b%22Tid%22%3a%225cdc5b43-d7be-4caa-8173-729e3b0a62d9%22%2c%22Oid%22%3a%221a8b9fa0-af57-4a1c-9390-22d1c201d622%22%7d




Best wishes,

Haoan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.sjtu.edu.cn/pipermail/adapt/attachments/20231121/ed884102/attachment.htm>