High-Dimension Human Value Representation in Large Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910893835550720 |
|---|---|
| author | Cahyawijaya, Samuel Chen, Delong Bang, Yejin Khalatbari, Leila Wilie, Bryan Ji, Ziwei Ishii, Etsuko Fung, Pascale |
| author_facet | Cahyawijaya, Samuel Chen, Delong Bang, Yejin Khalatbari, Leila Wilie, Bryan Ji, Ziwei Ishii, Etsuko Fung, Pascale |
| contents | The widespread application of LLMs across various tasks and fields has necessitated the alignment of these models with human values and preferences. Given various approaches of human value alignment, there is an urgent need to understand the scope and nature of human values injected into these LLMs before their deployment and adoption. We propose UniVaR, a high-dimensional neural representation of symbolic human value distributions in LLMs, orthogonal to model architecture and training data. This is a continuous and scalable representation, self-supervised from the value-relevant output of 8 LLMs and evaluated on 15 open-source and commercial LLMs. Through UniVaR, we visualize and explore how LLMs prioritize different values in 25 languages and cultures, shedding light on complex interplay between human values and language modeling. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_07900 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | High-Dimension Human Value Representation in Large Language Models Cahyawijaya, Samuel Chen, Delong Bang, Yejin Khalatbari, Leila Wilie, Bryan Ji, Ziwei Ishii, Etsuko Fung, Pascale Computation and Language Artificial Intelligence The widespread application of LLMs across various tasks and fields has necessitated the alignment of these models with human values and preferences. Given various approaches of human value alignment, there is an urgent need to understand the scope and nature of human values injected into these LLMs before their deployment and adoption. We propose UniVaR, a high-dimensional neural representation of symbolic human value distributions in LLMs, orthogonal to model architecture and training data. This is a continuous and scalable representation, self-supervised from the value-relevant output of 8 LLMs and evaluated on 15 open-source and commercial LLMs. Through UniVaR, we visualize and explore how LLMs prioritize different values in 25 languages and cultures, shedding light on complex interplay between human values and language modeling. |
| title | High-Dimension Human Value Representation in Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2404.07900 |