Inquiry about constrained MD in CP2K

[Jianzhi]

I am running a constrained MD simulation where atom A is driven to approach atom B, with atom B fixed during the simulation. The relevant parts of my input are as follows:

    &COLVAR
      &DISTANCE
        ATOMS 289 674
      &END DISTANCE
    &END COLVAR

......

  &CONSTRAINT
    &FIXED_ATOMS
      COMPONENTS_TO_FIX XYZ
      LIST 289
    &END FIXED_ATOMS
    &COLLECTIVE
      INTERMOLECULAR T
      COLVAR 1
      TARGET [angstrom] 31.9081
      TARGET_GROWTH [angstrom*fs^-1] -0.0008
    &END COLLECTIVE
    &LAGRANGE_MULTIPLIERS ON
       COMMON_ITERATION_LEVELS 1
       FILENAME ./lagrange.dat
    &END LAGRANGE_MULTIPLIERS
    SHAKE_TOLERANCE 0.00001
  &END CONSTRAINT

.......

Question 1: Why do the Shake Lagrangian multipliers appear with four components, for example:  

Shake  Lagrangian Multipliers:             0.188781938    0.000000000    0.000000000    0.000000000

In another test case I ran, only a single value was printed. What determines whether the output has one or multiple components?

Question 2: Are the units of the Shake Lagrangian multipliers given in Hartree/Bohr, regardless of the fact that the TARGET and TARGET_GROWTH values were specified in angstroms and angstrom/fs? I am asking because the free energy obtained from integrating the Shake Lagrangian multipliers appears to be unphysically large.