nyrikki

54 ms·

631.
by nyrikki 3y ago
Thank you, ordered it.
632.
by nyrikki 3y ago
In SR, you have the spacetime interval, it is more complex in GR. For constant velocity motion in flat spacetime: Δτ^2 Δs^2 = c^2 Δt^2-Δx^2-Δy^2-Δz^2 The spacetime interval is invariant, but notice it is also a scaler that combines space an
633.
by nyrikki 3y ago
Thanks for the suggestion. I personally used Dummit and Foote's book and found it useful, but like early Calculus with Spivak, it seems most people prefer clear and concise over slightly more comprehensive and rigorous while still bei
634.
by nyrikki 3y ago
The problem is that they aren't 'potato shaped' rotations, they are better thought of as hyperbolic rotations. Simultaneous distant events in Lorentz transformations will all lay on a shared hyperbola not a distorted circle.
635.
by nyrikki 3y ago
As someone who tutors adults, can you suggest a more digestible book for abstract algebra? While I was motivated, I used one of the typical college books. For me Abstract Algebra is what opened a lot of doors for me... but I am simply usin
636.
by nyrikki 3y ago
There is no objective concept of 'now' even at distance scales of Alpha Centauri with reasonable curvature and speeds. If you want to stick to special relativity, the Andromeda paradox will show the problem with simultaneity with
637.
by nyrikki 3y ago
That was once a popular belief, but we have moved past that historical concept. Here is a paper that is fairly accessible that may help. https://home.csulb.edu/~scrass/teaching/math456/articles/sen... It
638.
by nyrikki 3y ago
Ah the fun of known knowns, known unknowns, unknown unknowns, and weak negation. https://blog.research.google/2024/01/can-large-language-mode... Negation in the training data, exhaustion, or unknowable unknowns li
639.
by nyrikki 3y ago
This may be a bit pedantic, the nbody problem is not chaotic, it is harder, having riddled basins. > A riddled basin implies a kind of unpredictability, since exact initial data are required in order to determine whether the state of a s
640.
by nyrikki 3y ago
Yes please, the code quality on Devin was incredibly poor in all examples I traced down. At least from a maintainability perspective. I would like to see if this implementation is less destructive or at least more suitable for a red-green-r
641.
by nyrikki 3y ago
As a former storage architect, I should point out that defaults are compromises and it is far more complicated as any benchmarking is. Btrfs extent trees, XFS stride, stripe sizes. Partition alignment, WAL block and segment sizes all need t
642.
by nyrikki 3y ago
Due to the cliffhanger abstract, here is a part from the discussion that may help. > In our case, the manual curation of a proportion of triples revealed that Sherpa was able to extract more triples categorized as correct or partially co
643.
by nyrikki 3y ago
Impredicativity is the property you may want to dig into for formal proofs on why self references can be problematic. There is an important difference between semantically complete and syntactically complete that may cause some barriers. Gö
644.
by nyrikki 3y ago
Feed forward networks are effectively DAGs With the smaller model size, quantized models have less accuracy and less stable training, but large models take advantage of increased parallelism. Binary search is more linear, circuits are a bet
645.
by nyrikki 3y ago
PAC learning is compression. PAC learnable, Finite VC dimensionality, and the following form of compression are fully equivalent. https://arxiv.org/abs/1610.03592 Basically each individual neuron/perceptron just s
646.
by nyrikki 3y ago
In mining, the bean counters would use child and forced labor if they still could. But truck shovel routing is a version of VRP, trucks and shovels need to be matched or the approximations rise into the double digits of inefficiency. If you
647.
by nyrikki 3y ago
They chose the Caterpillar 797F, which is their largest offering. Not sure how they even think that the hoe that is loading them will also be able to load on-highway trucks. It would be like loading water glasses by tipping a 55gal barrel.
648.
by nyrikki 3y ago
> So, while various companies have set out to automate the giant dump trucks, having a continuous chain of smaller, self-driving trucks using standard parts and without the need for a driver cabin would be much more efficient So, to anyo
649.
by nyrikki 3y ago
FFT is easy fairly easy for transformers to find as a pattern. Note that OpenAI admits that pre-training (pattern finding and matching) is where the abilities on professional tests comes from. Even proofs in presburger arithmetic have an up
650.
by nyrikki 3y ago
Polysemous words, which sound the same and spell the same but have two or more than two different meanings are common. Each subfield defines rigourous terms and some may be overloaded. It is impossible to have any universal definition of wo
651.
by nyrikki 3y ago
It may not be known yet, and this project seems to be targeted at gaussian distributions, but wouldn't the simplicity bias reduce sensitivity? I mean attention in transformers works so well in part because OOD is typically close enoug
652.
by nyrikki 3y ago
As someone with a Gmail address that is my last name, I was signed up for Ashley Madison by a bot, and when I started getting the emails they wanted extortion money to delete it. The leak of that information had impact on my life. As Glass
653.
by nyrikki 3y ago
The oversimplified version is Laplac's deamon can't split quantum superposition. Superposition being inseparable is a large part on why the many words concept is popular with some people. It is about regaining a form of determini
654.
by nyrikki 3y ago
John von Neumann: "... the sciences do not try to explain, they hardly even try to interpret, they mainly make models. By a model is meant a mathematical construct which, with the addition of certain verbal interpretations, describes o
655.
by nyrikki 3y ago
Gravity in GR is a 'fictitious force' or an apparent force, meaning it is is an artifact of your chosen frame of reference. A feather falling the same speed as an elephant in the Earth reference frame is an example. It is the same
656.
by nyrikki 3y ago
While it is an open question, I found the paper that while not directly related is a reason to be concerned. Soft attention, applying a probabilistic curve across multiple neurons is why I think it is related. The Problem with Probabilistic
657.
by nyrikki 3y ago
Well xor is linearly inseparable, which is impossible with a single perceptron. > Our models by contrast make tractable predictions for what popu- lation and single-unit neural representations are required to support compositional genera
658.
by nyrikki 3y ago
>Tasks that are instructed using conditional clauses also require a simple form of deductive reasoning (if p then q else s) > Our models ofer several experimentally testable predictions outlining how linguistic information must be r
659.
by nyrikki 3y ago
>Yes, we don't really understand where emergent capabilities are coming from, at least not to extent of being able to predict them ahead of time ("if we feed it this amount of data, of this type, it'll learn to do X")
660.
by nyrikki 3y ago
I was at Sonosite in the mid 2000s, that was that entire company. Custom analog ASIC based, but portable tablets.

More ›