News for hackers. Stuff that matters.

alexmolas

alexmolas on Hacker News

Re:Jev Can't Be Calibrated on Thursday September 24, 2026 @02:49AM (#49827102)
Yes, good point. I meant that the Jev can't be calibrated for everyone out of the box. I find it important because this is one of their main claims. They literally say

> Outcomes assigned a probability of 0.2 should occur about 20% of the time.

and this is not true

Re:Jev Can't Be Calibrated on Wednesday September 23, 2026 @12:30PM (#49818597)
"Not especially good at classifying" does sound like a serious drawback for a classifier
Re:Jev Can't Be Calibrated on Wednesday September 23, 2026 @12:15PM (#49818345)
If you want calibrated probabilities you'll be forced to fine-tune it
Re:Jev Can't Be Calibrated on Wednesday September 23, 2026 @12:12PM (#49818284)
I don't know if it's reasonable. What it isn't is calibrated.
Re:Jev Can't Be Calibrated on Wednesday September 23, 2026 @11:16AM (#49817437)
But "problems about actual probabilities whose answers can't be known at all" are exactly the problems where calibration is important. Since calibration is one of the big claims about Jev I'd expect it to perform well in these problems.

Ask HN: What happened with Eureka Labs, Karpathy's startup?

Posted by alexmolas on Friday May 29, 2026 @05:07AM
from the ask-hn dept.

In 2024 Andrej Karpathy launched Eureka Labs [1], an AI-first education startup.Since the initial announcement, I haven't seen much public activity or updates, and the GitHub org [2] and course repo [3] appear empty or unused.Also, Recently Karpathy also joined Anthropic. He even removed Eureka Labs from his LinkedIn.Does anyone know what happened to Eureka Labs?[1]: https://eurekalabs.ai/ [2]:...

( Read More... | no comments | 3 points | on HN )
Re: on Thursday September 18, 2025 @05:25AM (#45287496)
Read the first paragraph of the Wikipedia page [1] for a lot of sources claiming the opposite.

[1]: https://en.wikipedia.org/wiki/Gaza_genocide

Older »