On a world like Venus, where the surface is unlivable , would it be possible for life to evolve in caves? And if so , is there any reason for intelligence to evolve in a subterranean organism?
Yes, if certain conditions were fulfilled.
First of all, I think a subterranean ocean is necessary. Or even better: a whole network of seas and lakes connected by tunnels. It would mean the crust of the planet is very thick and hard on the outside, but beneath there is a porous layer, probably full of soluble materials. The oceans would be very salty in comparison to Earth's oceans but that in fact may help you.
You would also need heat. If the planet is too hot on the surface, then it would be enough that the outside part of the crust is in certain places thin enough for the surface heat to warm the water. If not, we can go in the other direction, so deep that the waters are heated by the planet's core. Or by undersea vulcanoes. In fact, it would be nice to have some undersea volcanoes anyway.
Even on Earth there are certain ecosystems in places where the Sun doesn't shine and I don't mean it in the metaphorical sense. These microorganisms are called extremophiles, because they fare very well in conditions which are deadly for us, like very high temperatures and extreme salinity. On your planet they would probably be the first organisms to evolve, just next to undersea warm vents, and then they would slowly spread to colder areas, mixing with those evolving in other seas and lakes. But these conditions, good enough for microorganisms, seem to be too harsh for something more complex. For that you would need warm, but not too warm water and much lower salinity. Fortunately, that's possible as well if you give these extremophiles a few hundred million years.
That too happened on Earth in a way: Oxygen appeared first in the atmosphere as toxic waste of cyanobacteria feeding on CO2 (Great Oxygenation Event). In your case, a certain species of extremophiles would have to develop ability to turn salts into some kind of neutral material as a byproduct of their metabolic cycle. Gradually the not so hot waters away from undersea volcanoes would become less salty, and because of the abundance of extremophile microbiological life, they would become more rich in nutrients. And in such conditions more complex life could evolve.
As for intelligence... I guess with complex organisms the rest is just evolution working through millenia. Recently I have read a book about something like that, a civilization of "crustaceans" living on the seafloor on a planet covered by an ocean frozen on the surface: A Darkling Sea.
I'm not qualified in any field that relates to this question so feel free to kick this to the curb but my understanding is that the development of intelligence requires rapid changes in environment; rapid enough to require behavioural changes rather than being "solved" by natural selection. Being forced from one ecosystem into another is a common thread.
Hmm, https://www.quora.com/How-does-intelligence-arise suggests that the experts agree with me, in the broadest of senses. :-)
So I think you're looking for a subterrean ecology that is dramatically modified multiple times, forcing survivors to change behaviour, work together (developing socialisation and therefore language) and remember more information about good strategies and bad ones.
On that basis I think your barrier to development is that significant change in a subterrean ecology seems more difficult to escape so life forms either survive relatively unchanged or are completely wiped out, whereas on the surface there is more scope for survivors to relocate in a different ecosystem or be on the fringes of an event.
Underwater ecologies, I think, suffer from the same problem in reverse, in that it's so easy to move from one part of the ocean to another that it's difficult to "trap" creatures in a problem that they have to think or socialise their way out of. Aquatic intelligent life seems to have hit the "clever enough to handle any problem" limit although forcing an intelligent aquatic species to operate in an ecosystem with very little water (just enough for survival) might be an example of the kind of change that gave us intelligence.
So I guess you want a large, protected underground ecosystem that connects with other dramatically different ecosystems and then introduce events that force creatures to adapt. Settle, thrive and then re-disturb as many times as necessary.
So I'm agreeing with @makingthematrix about tunnels, seas and rivers but I don't think they're enough.