Not super important, but might be useful to have those logical operators in the core.
As a workaround implementation, they can be simulated with current functions. For example
-- t1 and t2 are ByteTensors with 0 or 1 only
function tensor_and(t1, t2)
return (t1 + t2):eq(2)
end
Implemented in #342.
Most helpful comment
Implemented in #342.