Natural Latents: Latent Variables Stable Across Ontologies
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909770211917824 |
|---|---|
| author | Wentworth, John Lorell, David |
| author_facet | Wentworth, John Lorell, David |
| contents | Suppose two Bayesian agents each learn a generative model of the same environment. We will assume the two have converged on the predictive distribution, i.e. distribution over some observables in the environment, but may have different generative models containing different latent variables. Under what conditions can one agent guarantee that their latents are a function of the other agents latents?
We give simple conditions under which such translation is guaranteed to be possible: the natural latent conditions. We also show that, absent further constraints, these are the most general conditions under which translatability is guaranteed. Crucially for practical application, our theorems are robust to approximation error in the natural latent conditions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_03780 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Natural Latents: Latent Variables Stable Across Ontologies Wentworth, John Lorell, David Probability Artificial Intelligence Information Theory Machine Learning Suppose two Bayesian agents each learn a generative model of the same environment. We will assume the two have converged on the predictive distribution, i.e. distribution over some observables in the environment, but may have different generative models containing different latent variables. Under what conditions can one agent guarantee that their latents are a function of the other agents latents? We give simple conditions under which such translation is guaranteed to be possible: the natural latent conditions. We also show that, absent further constraints, these are the most general conditions under which translatability is guaranteed. Crucially for practical application, our theorems are robust to approximation error in the natural latent conditions. |
| title | Natural Latents: Latent Variables Stable Across Ontologies |
| topic | Probability Artificial Intelligence Information Theory Machine Learning |
| url | https://arxiv.org/abs/2509.03780 |