Revealing the Parallel Multilingual Learning within Large Language Models
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_ | 1866908389898977280 |
|---|---|
| author | Mu, Yongyu Feng, Peinan Cao, Zhiquan Wu, Yuzhang Li, Bei Wang, Chenglong Xiao, Tong Song, Kai Liu, Tongran Zhang, Chunliang Zhu, Jingbo |
| author_facet | Mu, Yongyu Feng, Peinan Cao, Zhiquan Wu, Yuzhang Li, Bei Wang, Chenglong Xiao, Tong Song, Kai Liu, Tongran Zhang, Chunliang Zhu, Jingbo |
| contents | In this study, we reveal an in-context learning (ICL) capability of multilingual large language models (LLMs): by translating the input to several languages, we provide Parallel Input in Multiple Languages (PiM) to LLMs, which significantly enhances their comprehension abilities. To test this capability, we design extensive experiments encompassing 8 typical datasets, 7 languages and 8 state-of-the-art multilingual LLMs. Experimental results show that (1) incorporating more languages help PiM surpass the conventional ICL further; (2) even combining with the translations that are inferior to baseline performance can also help. Moreover, by examining the activated neurons in LLMs, we discover a counterintuitive but interesting phenomenon. Contrary to the common thought that PiM would activate more neurons than monolingual input to leverage knowledge learned from diverse languages, PiM actually inhibits neurons and promotes more precise neuron activation especially when more languages are added. This phenomenon aligns with the neuroscience insight about synaptic pruning, which removes less used neural connections, strengthens remainders, and then enhances brain intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_09073 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Revealing the Parallel Multilingual Learning within Large Language Models Mu, Yongyu Feng, Peinan Cao, Zhiquan Wu, Yuzhang Li, Bei Wang, Chenglong Xiao, Tong Song, Kai Liu, Tongran Zhang, Chunliang Zhu, Jingbo Computation and Language In this study, we reveal an in-context learning (ICL) capability of multilingual large language models (LLMs): by translating the input to several languages, we provide Parallel Input in Multiple Languages (PiM) to LLMs, which significantly enhances their comprehension abilities. To test this capability, we design extensive experiments encompassing 8 typical datasets, 7 languages and 8 state-of-the-art multilingual LLMs. Experimental results show that (1) incorporating more languages help PiM surpass the conventional ICL further; (2) even combining with the translations that are inferior to baseline performance can also help. Moreover, by examining the activated neurons in LLMs, we discover a counterintuitive but interesting phenomenon. Contrary to the common thought that PiM would activate more neurons than monolingual input to leverage knowledge learned from diverse languages, PiM actually inhibits neurons and promotes more precise neuron activation especially when more languages are added. This phenomenon aligns with the neuroscience insight about synaptic pruning, which removes less used neural connections, strengthens remainders, and then enhances brain intelligence. |
| title | Revealing the Parallel Multilingual Learning within Large Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2403.09073 |