Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2505.00204 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918005711044608 |
|---|---|
| author | Verma, Sumit Prasun, Pritam Jaiswal, Arpit Kumar, Pritish |
| author_facet | Verma, Sumit Prasun, Pritam Jaiswal, Arpit Kumar, Pritish |
| contents | As AI systems become embedded in real-world applications, ensuring they meet ethical standards is crucial. While existing AI ethics frameworks emphasize fairness, transparency, and accountability, they often lack actionable evaluation methods. This paper introduces a systematic approach using the Responsible AI Labs (RAIL) framework, which includes eight measurable dimensions to assess the normative behavior of large language models (LLMs). We apply this framework to Anthropic's "Values in the Wild" dataset, containing over 308,000 anonymized conversations with Claude and more than 3,000 annotated value expressions. Our study maps these values to RAIL dimensions, computes synthetic scores, and provides insights into the ethical behavior of LLMs in real-world use. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_00204 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RAIL in the Wild: Operationalizing Responsible AI Evaluation Using Anthropic's Value Dataset Verma, Sumit Prasun, Pritam Jaiswal, Arpit Kumar, Pritish Artificial Intelligence As AI systems become embedded in real-world applications, ensuring they meet ethical standards is crucial. While existing AI ethics frameworks emphasize fairness, transparency, and accountability, they often lack actionable evaluation methods. This paper introduces a systematic approach using the Responsible AI Labs (RAIL) framework, which includes eight measurable dimensions to assess the normative behavior of large language models (LLMs). We apply this framework to Anthropic's "Values in the Wild" dataset, containing over 308,000 anonymized conversations with Claude and more than 3,000 annotated value expressions. Our study maps these values to RAIL dimensions, computes synthetic scores, and provides insights into the ethical behavior of LLMs in real-world use. |
| title | RAIL in the Wild: Operationalizing Responsible AI Evaluation Using Anthropic's Value Dataset |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2505.00204 |