Your problem is probably not inside the tank, as there is only the 2 rods inside and no moving parts
. Currents involved in that loop are very small so, very easily messed by bad connection/corrosion. I could not test the gauge itself but what I did is proof the wiring
and gauge by connecting a potentiometer (variable resistance) at the sender end. If you can't do that, maybe try with an ohm meter.
What is the reading on gauge when tank full? Tank Empty? If the needle stay at zero, I would suspect a bad connection or bad wire. If reading over full, check screw connections of insulated rods. Screws must not pierce plastic, if so you will have too much current
flowing thru the loop.