A quantitative analysis of knowledge-learning preferences in large language models in molecular science
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912175723905024 |
|---|---|
| author | Liu, Pengfei Tao, Jun Ren, Zhixiang |
| author_facet | Liu, Pengfei Tao, Jun Ren, Zhixiang |
| contents | Deep learning has significantly advanced molecular modeling and design, enabling efficient understanding and discovery of novel molecules. In particular, large language models (LLMs) introduce a fresh research paradigm to tackle scientific problems from a natural language processing (NLP) perspective. LLMs significantly enhance our understanding and generation of molecules, often surpassing existing methods with their capabilities to decode and synthesize complex molecular patterns. However, two key issues remain: how to quantify the match between model and data modalities and how to identify the knowledge-learning preferences of models. To address these challenges, we propose a multi-modal benchmark, named ChEBI-20-MM, and perform 1263 experiments to assess the model's compatibility with data modalities and knowledge acquisition. Through the modal transition probability matrix, we provide insights into the most suitable modalities for tasks. Furthermore, we introduce a statistically interpretable approach to discover context-specific knowledge mapping by localized feature filtering. Our analysis offers an exploration of the learning mechanism and paves the way for advancing LLMs in molecular science. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_04119 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A quantitative analysis of knowledge-learning preferences in large language models in molecular science Liu, Pengfei Tao, Jun Ren, Zhixiang Machine Learning Computational Engineering, Finance, and Science Deep learning has significantly advanced molecular modeling and design, enabling efficient understanding and discovery of novel molecules. In particular, large language models (LLMs) introduce a fresh research paradigm to tackle scientific problems from a natural language processing (NLP) perspective. LLMs significantly enhance our understanding and generation of molecules, often surpassing existing methods with their capabilities to decode and synthesize complex molecular patterns. However, two key issues remain: how to quantify the match between model and data modalities and how to identify the knowledge-learning preferences of models. To address these challenges, we propose a multi-modal benchmark, named ChEBI-20-MM, and perform 1263 experiments to assess the model's compatibility with data modalities and knowledge acquisition. Through the modal transition probability matrix, we provide insights into the most suitable modalities for tasks. Furthermore, we introduce a statistically interpretable approach to discover context-specific knowledge mapping by localized feature filtering. Our analysis offers an exploration of the learning mechanism and paves the way for advancing LLMs in molecular science. |
| title | A quantitative analysis of knowledge-learning preferences in large language models in molecular science |
| topic | Machine Learning Computational Engineering, Finance, and Science |
| url | https://arxiv.org/abs/2402.04119 |