Towards Understanding Large Language Models for Multilingual Semantic Encoding

Session Number

Project ID: CMPS 23

Advisor(s)

Ermin Wei, Northwestern University

Discipline

Computer Science

Start Date

17-4-2024 9:40 AM

End Date

17-4-2024 9:55 AM

Abstract

Natural Language Processing (NLP) has witnessed significant advancements with the emergence of large language models (LLM) capable of understanding and generating human-like text. However, there remains a critical need to explore and understand their efficiency and effectiveness, especially in processing languages beyond English. This study aims to evaluate the efficiency of various large language models in capturing semantic meaning across English, German, and Spanish sentences. Principal Component Analysis (PCA) is utilized to identify important weights for understanding semantics. Through further experimentation with various sentence structures, we aim to identify factors contributing to the effectiveness of certain models. By pinpointing the strengths and weaknesses of different models, we aim to advance NLP research for multilingual applications.

Share

COinS
 
Apr 17th, 9:40 AM Apr 17th, 9:55 AM

Towards Understanding Large Language Models for Multilingual Semantic Encoding

Natural Language Processing (NLP) has witnessed significant advancements with the emergence of large language models (LLM) capable of understanding and generating human-like text. However, there remains a critical need to explore and understand their efficiency and effectiveness, especially in processing languages beyond English. This study aims to evaluate the efficiency of various large language models in capturing semantic meaning across English, German, and Spanish sentences. Principal Component Analysis (PCA) is utilized to identify important weights for understanding semantics. Through further experimentation with various sentence structures, we aim to identify factors contributing to the effectiveness of certain models. By pinpointing the strengths and weaknesses of different models, we aim to advance NLP research for multilingual applications.