Mason Yahr
@yahrmason.bsky.social
Does your company offer a roth option?
March 12, 2025 at 12:51 AM
Does your company offer a roth option?
Good on you man for staying above it. I take my match and spend all the rest
March 12, 2025 at 12:44 AM
Good on you man for staying above it. I take my match and spend all the rest
I am not going to lie, I am getting really sick of R. It’s way too permissible of downright heinous code.
February 26, 2025 at 3:50 AM
I am not going to lie, I am getting really sick of R. It’s way too permissible of downright heinous code.
Could use a facelift. yahrmason.github.io/bayes/gams-j...
Generalized Additive Models in Julia
An exploration of GAMs in Julia ‘from scratch’.
yahrmason.github.io
February 24, 2025 at 12:50 PM
Could use a facelift. yahrmason.github.io/bayes/gams-j...
I use Claude to explain formulas in academic papers to me. I have no idea if it’s correct, but it gives the appearance of doing a good job.
February 21, 2025 at 6:42 PM
I use Claude to explain formulas in academic papers to me. I have no idea if it’s correct, but it gives the appearance of doing a good job.
And there’s a surprising amount of information that simply isn’t sufficiently online to be surfaced from a nebulous “how do I analyze these data” search.
February 18, 2025 at 2:01 AM
And there’s a surprising amount of information that simply isn’t sufficiently online to be surfaced from a nebulous “how do I analyze these data” search.
Sorry, didn’t express my point well. He’s describing conservational search, and suggesting it is proper instruction. I think LLMs are fairly good for what he described he used it for, but he’s obviously extrapolating that way beyond reasonable utility.
February 18, 2025 at 1:43 AM
Sorry, didn’t express my point well. He’s describing conservational search, and suggesting it is proper instruction. I think LLMs are fairly good for what he described he used it for, but he’s obviously extrapolating that way beyond reasonable utility.
TBF what Cuban is talking about is more “conversational search” as opposed to actual instruction. I think LLMs are actually really great at this. But it’s basically just googling.
February 18, 2025 at 1:33 AM
TBF what Cuban is talking about is more “conversational search” as opposed to actual instruction. I think LLMs are actually really great at this. But it’s basically just googling.
I like this analogy. LLMs horrible tool for starting from 0. If you are already a hooper and you want to be more efficient in just grinding shots, then sure. Same if you are already a professional data person and you want to be more efficient at grinding boiler plate code etc
February 18, 2025 at 1:30 AM
I like this analogy. LLMs horrible tool for starting from 0. If you are already a hooper and you want to be more efficient in just grinding shots, then sure. Same if you are already a professional data person and you want to be more efficient at grinding boiler plate code etc
I use them a lot for guidance playing pokemon, as well as creative tasks such as coming up with clever nicknames for them. My favorite genre of use cases.
February 18, 2025 at 1:01 AM
I use them a lot for guidance playing pokemon, as well as creative tasks such as coming up with clever nicknames for them. My favorite genre of use cases.
Then the effects of climate change say “hold my beer”
February 16, 2025 at 1:06 AM
Then the effects of climate change say “hold my beer”
I would too. But I don’t think most employers agree. I couldn’t justify spending a Carroll’s worth of tuition on a PPE degree (though those were my favorite classes by far)
February 10, 2025 at 3:50 AM
I would too. But I don’t think most employers agree. I couldn’t justify spending a Carroll’s worth of tuition on a PPE degree (though those were my favorite classes by far)
A lot of the state is pretty red, but Madison is very much the opposite. It has a very good vibe about it.
January 25, 2025 at 2:37 PM
A lot of the state is pretty red, but Madison is very much the opposite. It has a very good vibe about it.
Outside of the cold (clearly big deal breaker), I think you would find Madison, WI pretty close to ideal.
January 25, 2025 at 2:34 PM
Outside of the cold (clearly big deal breaker), I think you would find Madison, WI pretty close to ideal.
Whatever the antithesis of “having a nice ring to it” is, that’s it.
January 14, 2025 at 3:48 AM
Whatever the antithesis of “having a nice ring to it” is, that’s it.
This was very interesting and informative, thanks for sharing.
January 14, 2025 at 12:52 AM
This was very interesting and informative, thanks for sharing.
“Fitting” if I am imposing some model structure and want to find the parameter set most consistent with the data. “Training” if I’m not, and I want the algorithm to “figure it out”.
But bc I know exactly what you mean… I use “fitting” because that’s how I want to be perceived
But bc I know exactly what you mean… I use “fitting” because that’s how I want to be perceived
January 14, 2025 at 12:49 AM
“Fitting” if I am imposing some model structure and want to find the parameter set most consistent with the data. “Training” if I’m not, and I want the algorithm to “figure it out”.
But bc I know exactly what you mean… I use “fitting” because that’s how I want to be perceived
But bc I know exactly what you mean… I use “fitting” because that’s how I want to be perceived
They are only 20 months, starting to get into the super fun age. They learn stuff incredibly fast. They can now name all the farm animals and tell you the sound they make.
January 14, 2025 at 12:06 AM
They are only 20 months, starting to get into the super fun age. They learn stuff incredibly fast. They can now name all the farm animals and tell you the sound they make.
