DANTE applies domain-adapted Vision Transformers, Multiple Instance Learning, and adaptive Dirichlet Process Mixture Models to unsupervisedly detect and triage extended transients in LIGO O4a data while stressing the need for native background recalibration.
However, the pipeline’s archi- tectural decoupling mathematically guarantees the ro- bustness of this finding without the need for extensive sensitivity simulations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
astro-ph.IM 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DANTE: A Reference-Guided Unsupervised Pipeline for Extended-Transient Anomaly Characterization in LIGO O4a
DANTE applies domain-adapted Vision Transformers, Multiple Instance Learning, and adaptive Dirichlet Process Mixture Models to unsupervisedly detect and triage extended transients in LIGO O4a data while stressing the need for native background recalibration.