Wrong multiplier (m) factor being saved by Spectre

Hello,

I'd like to save and get back the "m" parameter of my circuit devices in order to make some design counts I'm interested in. 

But, somehow the saved "m" value is always "1" in spite of the real value I assigned to it and Spectre actually used to run the simulation. I modify the multiplier "m" parameter either using the properties/CDF box in Virtuoso Schematic or manually editing the netlist.

To illustrate this, I just built a simple circuit with one transistor and run DC-sweep analysis with different values for "m". Its simplified netlist looks like this (I used montecarlo models in this one but it also happens with typical tt model):

#################################################################################

simulator lang=spectre
parameters mdut=2 vgn_par=0.8

include "/.../silterra/silterraC18G_cadenceIC6/CL016/models/spectre/c16_header.scs" section=c16_hp_monte

MNUT (vgn vgn 0 0) mnhp w=7u l=7u as=3.36p ad=3.36p ps=14.96u pd=14.96u \
        nrs=68.5714m nrd=68.5714m m=mdut sa=480n sb=480n sd=0 nf=1
VG (vgn 0) vsource dc=vgn_par type=dc

oppoint dc print=yes oppoint=rawfile
dc dc start=10m stop=2 param=vgn_par step=0.01 maxiters=150 maxsteps=10000 \
    oppoint=no hysteresis=no
...
#################################################################################

You can see the resultant waveforms in attached image, so it's clear that Spectre actually varies MNUT size using "m" and so its drain current does, but the saved "m" parameter is always "1". I wonder wheter this is related to either some Spectre simulation config or some PDK implementation issue.

I had a look to my PDK models and found out they state "m" like this:

########################################

section c16_hp_device

....

subckt nhp ( d g s b )

....

mint (d g s b) nhp  w=wf l=l sa=sa_comp sb=sb_comp ad=ad/nf pd=pd/nf as=as/nf ps=ps/nf nrs=nrs*nf nrd=nrd*nf m=(1*nf) delvto=delvto
ends nhp

subckt mnhp ( d g s b)

...

xmint (d g s b) nhp  w=wf l=l sa=sa_comp sb=sb_comp ad=ad/nf pd=pd/nf as=as/nf ps=ps/nf nrs=nrs*nf nrd=nrd*nf m=(1*nf)+(multn_hp * (aidn_hp/sqrt(w * l))) delvto=dvtn_hp * (avtn_hp/sqrt(w * l))
ends mnhp

########################################

Do you think that the statement "m=(1*nf)" has something to do with this?

Or maybe it is something related to Cadence save configurations?

Thank you in advance