Linkage
"I'm not good at math", car conversation with a 10-year-old girl about what math really is (illustrated, but with dubiously-reliable image hosting; G+)
Video on hypersphere packing, and the strange behavior of hyperspheres inside hypercubes
One of the six remaining candidates for the smallest odd k such that no number of the form k 2^n + 1 is prime has been knocked out (G+)
The Herschel enneahedron, a nice symmetric geometric realization of the simplest non-Hamiltonian polyhedron
Antiparallelogram-based linkage traces out a three-lobed curve (more linkage curves linked from G+)
Secret deal for funneling UK university money to Elsevier (G+)
On the number of ordinary lines determined by sets in complex space
Comments:
2016-12-01T21:13:57Z
That packing thing is interesting. One ML professor in the draft of his book claimed that the sphere in the center of the cube touching four other spheres became a "porcupine" in high dimensions: it sticks through the wholes and outside the cube. That made me scratch my head, because a convex round object should be convex and round in any projection. However, I have figured out that the sphere becomes so big that it starts sticking outside (actually this is clearly backed up with simple math). But it still has to be round, so it won't look like a porcupine (which is not convex).
Apparently, this sticking phenomenon is possible due to the space emptiness. The four other spheres occupy a tiny fraction of the cube space, so, apparently, they do not prevent the central sphere from going (as dim increases). And eventually it can even stick out of the cube! This seems to be an extremely interesting packing paradox.