Presentation #100.05 in the session AGN.
An ongoing campaign of Swift-XRT and -UVOT observations has discovered hundreds of X-ray/UV/optical counterparts to low-flux gamma-ray sources that were previously unassociated in Fermi catalogs, lacking known counterparts. Applying machine learning classifiers like random forests and neural networks, we found that many of the counterparts are likely blazars or pulsars in similar proportion to the brighter gamma-ray blazars and pulsars of established catalogs. Investigating these low-luminosity blazars identified in the unassociated catalog can test predictions of theories like the blazar sequence and add to established samples by reducing flux-limited biases. With Fermi gamma-ray data, X-ray/UV/optical observations from Swift, and archival WISE and radio fluxes, we construct broadband SEDs for the newly identified blazars. We estimate synchrotron and high-energy peak frequencies and Compton dominance for the newly identified blazars, showing that most are similar to high-synchrotron-peak BL Lac objects, in agreement with predictions of the blazar sequence framework. Finally, we fit the SEDs of the blazars with a synchrotron-self-Compton jet emission model, obtaining estimates for physical jet parameters. These results allow for direct comparisons between bright blazar catalogues and new, low-luminosity blazars from the unassociated list.