patrick0d

6 ms·

1.
by patrick0d 24d ago
thank you from sharing negative results! this was something that was missing from the general competition I think
2.
by patrick0d 2mo ago
Inspired by parameter golf and speedrun approaches I make the case for picking loss functions like a wallclock for LoRA on AI safety targets. The result when I tried it was a functional distillation of an Sparse AutoEncoder into a 5.3MB pro
3.
by patrick0d 11mo ago
paper looks nice! i think what they found was that they can recover the input sequence by trying all tokens from the vocab and finding a unique state. they do a forward pass to check each possible token at a given depth. i think this is sin
4.
by patrick0d 1y ago
anyone who spends 10 years writing performance reviews and rebuilding a service they already rebuilt last year, all to collect RSUs, was never going to land us on mars
5.
by patrick0d 1y ago
Well don't do it and instead of using an off the shelf library that is known to work while the rest of the development team isn't reinventing the wheel. Doing it for fun and education is fine of course.
6.
The AI Safety Puzzle Everyone Avoids: How to Measure Impact, Not Intent(lesswrong.com)
1 points by patrick0d 1y ago|1 comments
7.
by patrick0d 1y ago
I am an AI interpretability researcher and have a new proposal for a way to measure the per token contribution of each head and neuron in LLMs. I found that the normalisation that happens in every LLM is avoided by modern attribution metho
8.
by patrick0d 2y ago
Maybe eating honey has never happened in the history of human nature, and it's fool-hardy.
9.
by patrick0d 2y ago
the average human certainly can not run that fast. treadmills in gyms usually cap at 20km/h and how many people on the street do you think could handle it? (very few of course)
10.
by patrick0d 2y ago
this vibe check is more insightful to me than the popular evals. nice job!
11.
by patrick0d 2y ago
they could have made the law: >if you collect users data >you must ask first >add a yes or no button on a banner so they can pick but instead the eu citizens were let down by the legislators
12.
by patrick0d 2y ago
this should be the real headline. I am certain and beyond any doubt that I speak for all of us hackernews readers when I say that I have no frame of reference on the value of cryptocurrencies without comparing first with the Malaysian Ringg
13.
by patrick0d 2y ago
What's your definition?
14.
by patrick0d 2y ago
All you need to do to see some change is pick a goal and commit a long time towards it. So if you have a long time maybe you can help with this project that you would like to see.
15.
by patrick0d 3y ago
I said a good teenager might run 10.7 and you have found a list of teenagers (I don't know how old they are) that can do just that. Very few people sprint after they leave college but if they spent time becoming stronger and more physi
16.
by patrick0d 3y ago
It is a good time and I gave the example of a teenager since they will likely be less physically developed, weaker and less experienced than a pro. In a place where sprinting is popular like in the US 10.70 is a pretty achievable for someon
17.
by patrick0d 3y ago
>height and weight, fast- and slow-twitch muscle mass, cardiovascular conditioning, flexibility and elasticity, and probably more. When we try to figure out how to measure someones speed (say to compare Usain Bolt to another high quality
18.
by patrick0d 3y ago
Those people who were suffering from curable blindness that had been forgotten by society sure are marketed, businessed and entertained by their restored sight. He's doing good things to improve peoples lives and entertaining millions
19.
by patrick0d 3y ago
The worry is not that chatgpt will take over the world. It is that a future system will be unaligned with human interests and once it is created by gradient descent (the internals of the system are not understood by anyone - they're ju
20.
by patrick0d 4y ago
There is a book called "The perfect machine" about building this observatory.I found it very interesting since you might think it's not as hard as it is to make a piece of glass a few meters across and make it reflective.
21.
by patrick0d 4y ago
The parts of chemistry that work are just rebrandings of physics and the parts of physics that work are just rebrandings of math. Nanotech is a subset of physics with connections to solid state physics, quantum mechanics, materials science,
22.
by patrick0d 4y ago
Why not take the 100 billion from somewhere other than frontier physics research. I always consider these articles small minded since the people writing them don't consider projects outside of their domain that eat up more funding than
23.
by patrick0d 5y ago
He's referring to Carlsen's prodigious fantasy football skill.
24.
by patrick0d 5y ago
We should add treads and a cannon while we're at it. Is outer space counted when we say "all terrain"?
25.
by patrick0d 5y ago
Eddington said: "The law that entropy always increases—the Second Law of Thermodynamics—holds, I think, the supreme position among the laws of Nature. If someone points out to you that your pet theory of the universe is in disagreement
26.
by patrick0d 5y ago
Because the universe is expanding the older and farther away light is the redder it looks to us. So by collecting this infrared light we can see farther back in time than we could with any other instrument before.
27.
by patrick0d 5y ago
There's a book called "Skunk Works" by Ben Rich about the making of different spy planes. There's some mentions in the book about satellites and what existed at the time how projects were carried out.