Euclid is delivering millions of galaxy images, far too many for humans to classify by hand. A new approach uses sparse autoencoders to uncover the hidden morphological features inside deep learning models, revealing both familiar structures and entirely new ones.
Clumpy galaxies dominate the early universe, yet their local counterparts are hard to find. A new machine‑learning approach learns to spot these hidden clumps and opens the door to studying them in far greater detail.
Citizen scientists can make important contributions to the study of galaxies (like the discovery of new gravitational lenses!), and their work can also be used to train better machine learning models.
AI is everywhere, whether we want it or not. In climate modeling, AI integration is improving simulations, which is good, because the environmental impact of AI is going to change our climate even faster.
How do you search through 99.6 million images for weird and interesting galaxies? Instead of wasting years of your life doing it manually, you can use AnomalyMatch.
Is AI the real expert on galaxy morphology evolution? Read today’s bite to find out!