Value-Action Alignment in Large Language Models under Privacy-Prosocial Conflict
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912806435028992 |
|---|---|
| author | Chen, Guanyu Yu, Chenxiao Hu, Xiyang |
| author_facet | Chen, Guanyu Yu, Chenxiao Hu, Xiyang |
| contents | Large language models (LLMs) are increasingly used to simulate decision-making tasks involving personal data sharing, where privacy concerns and prosocial motivations can push choices in opposite directions. Existing evaluations often measure privacy-related attitudes or sharing intentions in isolation, which makes it difficult to determine whether a model's expressed values jointly predict its downstream data-sharing actions as in real human behaviors. We introduce a context-based assessment protocol that sequentially administers standardized questionnaires for privacy attitudes, prosocialness, and acceptance of data sharing within a bounded, history-carrying session. To evaluate value-action alignments under competing attitudes, we use multi-group structural equation modeling (MGSEM) to identify relations from privacy concerns and prosocialness to data sharing. We propose Value-Action Alignment Rate (VAAR), a human-referenced directional agreement metric that aggregates path-level evidence for expected signs. Across multiple LLMs, we observe stable but model-specific Privacy-PSA-AoDS profiles, and substantial heterogeneity in value-action alignment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_03546 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Value-Action Alignment in Large Language Models under Privacy-Prosocial Conflict Chen, Guanyu Yu, Chenxiao Hu, Xiyang Computation and Language Artificial Intelligence Human-Computer Interaction Machine Learning Large language models (LLMs) are increasingly used to simulate decision-making tasks involving personal data sharing, where privacy concerns and prosocial motivations can push choices in opposite directions. Existing evaluations often measure privacy-related attitudes or sharing intentions in isolation, which makes it difficult to determine whether a model's expressed values jointly predict its downstream data-sharing actions as in real human behaviors. We introduce a context-based assessment protocol that sequentially administers standardized questionnaires for privacy attitudes, prosocialness, and acceptance of data sharing within a bounded, history-carrying session. To evaluate value-action alignments under competing attitudes, we use multi-group structural equation modeling (MGSEM) to identify relations from privacy concerns and prosocialness to data sharing. We propose Value-Action Alignment Rate (VAAR), a human-referenced directional agreement metric that aggregates path-level evidence for expected signs. Across multiple LLMs, we observe stable but model-specific Privacy-PSA-AoDS profiles, and substantial heterogeneity in value-action alignment. |
| title | Value-Action Alignment in Large Language Models under Privacy-Prosocial Conflict |
| topic | Computation and Language Artificial Intelligence Human-Computer Interaction Machine Learning |
| url | https://arxiv.org/abs/2601.03546 |