Identifying Metric Structures of Deep Latent Variable Models
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_ | 1866908384608911360 |
|---|---|
| author | Syrota, Stas Zainchkovskyy, Yevgen Xi, Johnny Bloem-Reddy, Benjamin Hauberg, Søren |
| author_facet | Syrota, Stas Zainchkovskyy, Yevgen Xi, Johnny Bloem-Reddy, Benjamin Hauberg, Søren |
| contents | Deep latent variable models learn condensed representations of data that, hopefully, reflect the inner workings of the studied phenomena. Unfortunately, these latent representations are not statistically identifiable, meaning they cannot be uniquely determined. Domain experts, therefore, need to tread carefully when interpreting these. Current solutions limit the lack of identifiability through additional constraints on the latent variable model, e.g. by requiring labeled training data, or by restricting the expressivity of the model. We change the goal: instead of identifying the latent variables, we identify relationships between them such as meaningful distances, angles, and volumes. We prove this is feasible under very mild model conditions and without additional labeled data. We empirically demonstrate that our theory results in more reliable latent distances, offering a principled path forward in extracting trustworthy conclusions from deep latent variable models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_13757 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Identifying Metric Structures of Deep Latent Variable Models Syrota, Stas Zainchkovskyy, Yevgen Xi, Johnny Bloem-Reddy, Benjamin Hauberg, Søren Machine Learning Deep latent variable models learn condensed representations of data that, hopefully, reflect the inner workings of the studied phenomena. Unfortunately, these latent representations are not statistically identifiable, meaning they cannot be uniquely determined. Domain experts, therefore, need to tread carefully when interpreting these. Current solutions limit the lack of identifiability through additional constraints on the latent variable model, e.g. by requiring labeled training data, or by restricting the expressivity of the model. We change the goal: instead of identifying the latent variables, we identify relationships between them such as meaningful distances, angles, and volumes. We prove this is feasible under very mild model conditions and without additional labeled data. We empirically demonstrate that our theory results in more reliable latent distances, offering a principled path forward in extracting trustworthy conclusions from deep latent variable models. |
| title | Identifying Metric Structures of Deep Latent Variable Models |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2502.13757 |