Observing the progenitor of core-collpase supernova before their demise can often shed light of the final moments of a massive stars life. However, observations of one class of transients bring to question our understanding of how exactly a star may die, and if in fact it dies at all.

AT 2016jbu is an interacting transient, with similarities to Type IIn supernovae. AT 2016jbu displays a history of erratic outbursts, lasting for at least 10 years prior to two luminous explosions, that latter of which reaches supernova-like magnitudes. The circumstellar environment of AT 2016jbu is highly asymmetrical, displaying multi peaked emission lines after maximum brightness. The historical light curve of AT 2016jbu is reminiscent of the supernova impostor, Eta carinae, prior to the great eruption in the 1840s. At late times, AT 2016jbu displays tentative signs of core-collpase. In particular the lack of nebular emission lines, and a low inferred mass of radioactive nickel. Progenitor observations reveal that AT2016jbu has exploding prematurely. We find the progenitor has the appearance of a warm 20 solar mass, yellow hypergiant, rather the expected red supergiant or wolf-rayet star. The questions thus stands, are we witnessing the death of a star or some non-terminal explosion?

Using the Hubble Space Telescope, we return to the site of AT 2016jbu almost 5 years post explosion to search for a surviving star. While identifying a source at the transient site, we find the visual brightness has changed by more than 2 mags. We postulate that a surviving source has been enshrouded by dust produced in the 2016 explosion, and find an unrealistic amount is needed for a non-terminal eruption.

With this, we present a possible progenitor scenario, involving a binary system, and discuss the possibility of observing the binary companion in the future.