Consider the optimization problem: min x_1^2 + x_2^2 - 6x_1 - 4x_2 + 3 subject to the constraints: x_1 + x_2 >=1, x_2<=1, -x_1+x_2 >= -1. Suppose that we are running projected gradient descent, and at time t, we are at the point x_1(t) = 1/2, x_2(t) = 1/2. After a PGD step with eta = 1/5, we are at the point