Most volume renderers lack a good transfer function editor.
When analyzing volumes, especially explorative analysis, the most effective tool is to dial in colors and opacities for certain value ranges, in order to find structures.
The volume rendering engine I have been working on uses a histogram for the value distribution, and on top of it, one can draw lines that indicate the opacity. Additionally, one can set colors to the control points, which are then linearly interpolated for the given ranges.
Not true. You need a distance matrix (for classical PCA it's a covariance matrix), which scales quadratically with the number of points you want to compare.
If you have 1 Mio. vectors, each creating a float entry in the matrix, you will end up with approx (10^6)^2 / 2 unique values, which is roughly 2000Gb of memory.
Some architectures are relatively well understood. Eg in CNNs, the first layers detect low level features like edges, gradients, etc. The next layer then combines these features to more complex structures like corners or circles. Next layer will combine these features to even higher level features and so on. [1]
Typically, you can take a pre-trained model and retrain it on your new dataset by only changing the weights of the last layer(s).
Some loss functions even measures the difference between the high-level features of two images, typically extracted from a pre-trained CNN (Perceptual Loss).
I have not seen many work on explainable AI regarding large language models. I remember many very nice visualizations and visual analysis tools trying to comprehend, what the network „is seeing“ (eg. in the realm of image classification) or doing
Definetly not true. This can be the best password in certain conditions.
You should not put your "good" passwords to any shady site out there. You have no idea how passwords are stored on all these platforms one is registering to.
If you can live with the fact, that an account may be hacked, then go for a super easy password if you want.
Having no framework comes with having no documentation.
If you select a framework and work in the same scope as the framework intends to, then you can onboard new engineers much more easily by telling them: We are working as the framework defines, look into the documentation and you will understand the basic architecture of our software.
Of cause, devs can maintain their own documentation, but I guess this usually is something that is not done very well.
My personal experience: As soon as a company started their software without a framework, it became a huge mess. While onboarding, I observed very bad architectural decisions and even very severe security issues.
I'd say, only go with no framework, if you are experienced enough to technically create an own (good) framework.
From my observations, going without a framework was usually a decision done by young engineers.
All your advantages of Java also apply to Python. Why would Java be the better choice compared to Python? Or to rephrase this question, why do you think that Python is only good for small teams (with an eye on all these massive billion dollar projects like Instagram which were built in python)?