Epistemic Uncertainty Quantification for Pre-trained VLMs via Riemannian Flow Matching
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916031549669376 |
|---|---|
| author | Ju, Li Nautiyal, Mayank Hellander, Andreas Vats, Ekta Singh, Prashant |
| author_facet | Ju, Li Nautiyal, Mayank Hellander, Andreas Vats, Ekta Singh, Prashant |
| contents | Vision-Language Models (VLMs) are typically deterministic in nature and lack intrinsic mechanisms to quantify epistemic uncertainty, which reflects the model's lack of knowledge or ignorance of its own representations. We theoretically motivate negative log-density of an embedding as a proxy for the epistemic uncertainty, where low-density regions signify model ignorance. The proposed method REPVLM computes the probability density on the hyperspherical manifold of the VLM embeddings using Riemannian Flow Matching. We empirically demonstrate that REPVLM achieves near-perfect correlation between uncertainty and prediction error, significantly outperforming existing baselines. Beyond classification, we also demonstrate that the model also provides a scalable metric for out-of-distribution detection and automated data curation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_21662 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Epistemic Uncertainty Quantification for Pre-trained VLMs via Riemannian Flow Matching Ju, Li Nautiyal, Mayank Hellander, Andreas Vats, Ekta Singh, Prashant Machine Learning Vision-Language Models (VLMs) are typically deterministic in nature and lack intrinsic mechanisms to quantify epistemic uncertainty, which reflects the model's lack of knowledge or ignorance of its own representations. We theoretically motivate negative log-density of an embedding as a proxy for the epistemic uncertainty, where low-density regions signify model ignorance. The proposed method REPVLM computes the probability density on the hyperspherical manifold of the VLM embeddings using Riemannian Flow Matching. We empirically demonstrate that REPVLM achieves near-perfect correlation between uncertainty and prediction error, significantly outperforming existing baselines. Beyond classification, we also demonstrate that the model also provides a scalable metric for out-of-distribution detection and automated data curation. |
| title | Epistemic Uncertainty Quantification for Pre-trained VLMs via Riemannian Flow Matching |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.21662 |