Is it possible for a creature to change forms by turning into a liquid then a solid?

Inside the chrysalis a caterpillar turns into goo, which then becomes a moth. Many parts of its body break down and new ones are made in different shapes. However it does not become completely liquid, and it doesn't do much while it's gooey.

A floppy but highly structured body like an octopus can think and move (though a hard body is stronger; our bones give our muscles better leverage). A liquid body will be very stupid and not very capable.

Your creature will not be able to make fast or complex or powerful movements if its body has no structure. It will not be able to think if its nervous system has no structure. If it had almost no structure at all, it would have the intelligence and strength of a slime mold.

After the change, it will not remember its previous life unless the nervous system retained some structure during the transition. A moth can react to stimulus it learned about as a caterpillar, because its nervous system did not lose all its connections while the body changed shape.

No

Well, 'no' seems like a really strong answer because I'm saying such a creature can't happen, ever, but I'm pretty comfortable saying that, given a 'rodent-sized' creature. There are just too many problems involved with a creature turning into a liquid.

Objects which are solid don't typically become liquid of their own volition. There are traditionally three ways for a solid to become a liquid. The first is the temperature increases, i.e. ice to water. This doesn't work for a living organism, because all most organic processes use proteins and catalysts and those proteins and catalysts will denature and be destroyed outside a certain range of temperature. A change of temperature which is large enough to melt the solid object which is the animal into a liquid will certainly denature all these very fragile proteins. Dead animal. The second method is a change in pressure can occasionally change objects from solid to liquid, though the solid needs to be relatively close to the melting point anyway and the pressure change if a severe lack of pressure. This also isn't very survivable and is entirely reliant on an external process.

The third, and last method, is changing the state of the object. For instance, fat is a solid at room temperature, however, should you make a slight change to the chemical structure of fat (a double-bond within a fatty acid change) and then it becomes oil - a liquid at room temperature. While this is a valid method for turning into a liquid, there are two very significant problems with doing this to an animal - energy costs and reconstruction. You would need to spend energy to turn the creature's entire structure into a new liquid-capable one which would take an absurd amount of energy to accomplish and you would need to have an even larger expenditure of energy to turn it back into a solid. Additionally, you would need to somehow keep all the necessary components for the very specific biological functions within the same cells - for instance, you'd need to be able to keep the heart functioning when its reconstructed and not have the heart cells just go floating around the liquid animal.

It may be possible when dealing with very small organisms on the cellular level, but not an animal like a rodent.