Candle cheatsheet needs more examples.

The cheatsheet in the readme really needs more examples...

Basic math Tensor + u32 for instance is super simple in python but... ya
tensor[:, None, None, :] lookups...
how to handle [-1] lookups...
best way to update slices with assign_slice, that can be done with a very simple expression in python...

Start with the cheatsheet in the README and review the existing examples and formatting. Add concise examples covering Tensor plus u32, tensor[:, None, None, :], [-1] lookups, and assign_slice updates; done means these requested operations are documented clearly in the cheatsheet.

Written by the indexing model from the issue text.