StarXiv: a podcast discussing the latest astronomy papers
Jun 8, 2026 · 36 min · 11 segments
In this episode, Michelle and Nicole discuss machine learning methods to classify merging galaxies and different classes of supernova. They also look into whether planetary engulfment can explain…
No entities detected.
Ooh
I might only cry a little bit.
[laughs]
[laughs]
But, uh, yeah, but two d- very different science topics at least, and, and nothing like what you're doing probably.
Phew.
[laughs]
Uh, so my first paper is called Vision Language Models, or VLMs, as Ensem- Ensembles Achieve Human Expert Accuracy for Galaxy Merger Classifications.
And one thing that actually drew me to the abstract of this one is that, you know you have the notes section at the bottom of the archive abstract-
Mm
... as people are like, "14 pages, X number of figures, key figure is blah." This one actually talked about how, you know, they used AI as a research assistant in terms of sort of helping with, like, idea generation coding and draft under human supervision, like post-human super-
Oh.
I was like, "Ah, like a nice explicit statement right up front in the abstract-
Yeah
Ooh
I might only cry a little bit.
[laughs]
[laughs]
But, uh, yeah, but two d- very different science topics at least, and, and nothing like what you're doing probably.
Phew.
[laughs]
Uh, so my first paper is called Vision Language Models, or VLMs, as Ensem- Ensembles Achieve Human Expert Accuracy for Galaxy Merger Classifications.
And one thing that actually drew me to the abstract of this one is that, you know you have the notes section at the bottom of the archive abstract-
Mm
... as people are like, "14 pages, X number of figures, key figure is blah." This one actually talked about how, you know, they used AI as a research assistant in terms of sort of helping with, like, idea generation coding and draft under human supervision, like post-human super-
Oh.
I was like, "Ah, like a nice explicit statement right up front in the abstract-
Yeah
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.