markisus

10 ms·

91.
by markisus 2y ago
Article says it’s about stack addresses but I think it should apply more generally to dynamically allocated memory as well.
92.
by markisus 2y ago
Here is his 1984 article published in PRL in which he highlights that cellular automata can be realized as physical systems and therefore computational intractabilities will manifest in models of physics. https://content.wolfram.
93.
by markisus 2y ago
It's defined on page 5. > In this paper we are considering the behaviour of ∆(HΛ(L)) in the thermodynamic limit, that is, when L → ∞ So it's the same as the regular mathematical limit. There must be some thermodynamic implicati
94.
by markisus 2y ago
Here is a free updated version from 2022. https://arxiv.org/pdf/1502.04573 Edit: Alright, I'm back after skimming the paper. This is a halting oracle. Here is what I understand, with my limited knowledge of QM.
95.
by markisus 2y ago
Regarding faulty intermediate versions, I think that’s the point. The diffusion process can correct wrong tokens with the global state implies it.
96.
by markisus 2y ago
Surprising. What was the query?
97.
by markisus 2y ago
I think you may be right. I just looked up Kinect on eBay and you can get one for twenty bucks.
98.
by markisus 2y ago
I wonder why the Kinect is still so popular given the many more recent projects that have come to market, eg Realsense, Zed, Orbec. I’ve recently been using the Realsense for robotics and they are satisfactory (although they have been drast
99.
by markisus 2y ago
Is the curriculum that different? The fundamentals are the same no matter where you go.
100.
by markisus 2y ago
It's a little hard to understand your criticism. From just reading the original article, I had assumed that Claude could decompile binaries. The author said > Understand dear reader that this technique can be done on any programming
101.
by markisus 2y ago
Very interesting! Do you know if there were any studies about whether this improves performance?
102.
by markisus 2y ago
The article contains a reference to a much more impressive task where a user automatically decompiled a binary exe game into Python. But I read their original post and here is what that user said. > Several critics seemed to assume I cla
103.
by markisus 2y ago
That's true in the realm of LLMs. But even in this case, the position information is added only into the first layer. Tokens in later layers can choose to "forget" this information. In addition there are applications of trans
104.
by markisus 2y ago
I was careless in my thinking, thanks for the correction. I was imagining since you sum the group elements in any order, there is a permutation invariance. But the group elements themselves play the role of the "token index" and g
105.
by markisus 2y ago
I’m thinking the integers mod 2^n where n is something computers are good at (8, 32, 64). You have hardware support the group operation.
106.
by markisus 2y ago
The Fourier transform is taken along the “token” dimension. However, in many applications, this dimension is not meaningful. That’s why transformers are a great option for consuming data which is permutation invariant. I would like to see a
107.
by markisus 2y ago
Yes, by advanced undergraduate, I meant very advanced undergraduate. But when I was in undergrad I always heard about some students like this who were off in the graduate classes. And then in grad school, there was even a high school stud
108.
by markisus 2y ago
Here is a corresponding introduction I found very useful, for readers with advanced undergraduate / graduate level math knowledge. https://almostsuremath.com/stochastic-calculus/
109.
by markisus 2y ago
The interesting thing about this project is that it shouldn’t even be possible if the chip behaved as an abstract logical circuit since then it would simply implement a finite automation. You must abuse the underlying physics to make the lo
110.
by markisus 2y ago
This is basically Gaussian splat using cubes instead of Gaussians. The cube centers and sizes choices are discrete and non overlapping, hence the name “sparse voxel”. The qualitative results and rendering speeds are similar to Gaussian spla
111.
by markisus 2y ago
It’s interesting that the conjecture was proven via case by case analysis, with each case demanding different techniques. It’s almost a coincidence that all finite groups have this property, since each group has the property because of a di
112.
by markisus 2y ago
It could be like how people enjoy assembling Gundam models. Sometimes it’s fun just watching something beautiful emerge from many separate components. People also enjoy the “how stuff is made” type videos. Simulating a manufacturing process
113.
by markisus 2y ago
I like this idea. The machine shop can be quite intimidating. It could be a fun little experience to get people exposed to heavy equipment in a safe environment.
114.
by markisus 2y ago
An interesting essay on the history and precursors to Dedekind’s construction of the Reals. I was disappointed that the author didn’t talk about some of the modern opponents of the “real numbers.” For example I heard this interesting podcas
115.
by markisus 2y ago
If by helping anyone you mean helping the math challenged (which is technically almost everyone) then I would be inclined to agree. But to the quant crowd I’m guessing that couching diffusion in the language of stochastic calculus is helpfu
116.
by markisus 2y ago
I like the glorified mitosis analogy. And psychology is glorified particle physics. It’s kind of true, but you just know there is something else going on there that might explain it better.
117.
by markisus 2y ago
The r1 somehow knew at an early stage that the message was HELLO but it couldn’t figure out the reason. Even at the end, its last “thought” insists that there is an encoding mistake somewhere. However the final message is correct. I wonder
118.
by markisus 2y ago
This is a great point. I remember my first computer programming class was Java. The teacher said “just type public static void main( string[] args) at the top”. I asked why and he said it didn’t matter for now just memorize that part. That
119.
by markisus 2y ago
I think I'm mostly with you except a couple things. > Are the scientists interacting? So say Alice and Bob interact, forming an event in RQM. Then they both separately interact with Charlie, generating two new events. These two new
120.
by markisus 2y ago
RQM is an interesting viewpoint on QM that I have not encountered before, even though it was apparently introduced in 1996. It says that the physical world is a collection of discrete events, doing away with continuous time. It also says th

More ›