Running and Tensor Constraints


[Brown et al Fig 16]

Modest improvement - running favored at approx 2 sigma