We are able to put an excellent determine on how a lot we all know concerning the universe: 5%. That’s how a lot of what’s floating about within the cosmos is extraordinary matter—planets and stars and galaxies and the mud and fuel between them. The opposite 95% is darkish matter and darkish power, two mysterious entities aptly named for our incapability to make clear their true nature.
Earlier work has begun pulling aside these dueling forces, however darkish matter and darkish power stay shrouded in a blanket of questions—critically, what precisely are they?
Enter the Vera C. Rubin Observatory, one in all our 10 breakthrough applied sciences for 2025. Boasting the biggest digital digicam ever created, Rubin is anticipated to check the cosmos within the highest decision but as soon as it begins observations later this 12 months. And with a greater window on the cosmic battle between darkish matter and darkish power, Rubin would possibly slim down present theories on what they’re made from. Right here’s a have a look at how.
—Jenna Ahart
This story is a part of MIT Know-how Overview Explains, our sequence untangling the advanced, messy world of expertise that will help you perceive what’s coming subsequent. You possibly can learn extra from the sequence right here.
Anthropic has a brand new method to defend giant language fashions towards jailbreaks
What’s new? AI agency Anthropic has developed a brand new line of protection towards a typical form of assault referred to as a jailbreak. A jailbreak tips giant language fashions (LLMs) into doing one thing they’ve been skilled to not, akin to assist someone create a weapon. And Anthropic’s new method could possibly be the strongest protect towards the assaults but.
How they did it: Jailbreaks are a form of adversarial assault: enter handed to a mannequin that makes it produce an surprising output. Regardless of a decade of analysis there may be nonetheless no method to construct a mannequin that isn’t weak. However, as an alternative of attempting to repair its fashions, Anthropic has developed a barrier that stops tried jailbreaks from getting by way of and undesirable responses from the mannequin getting out. Learn the complete story.
