Something to keep in mind - and this is my own weird musings, mind.

I grew up around horses; and have certain opinions about the considerable intelligence of these magnificent creatures. Also; I spent considerable time with a dolphin researcher (OK; I was dating her.)

Researchers studying animal intelligence have a tendancy to do so by comparing the animal's intelligence with our own. In other words; they tend to make assumptions of a being's intelligence based on what Human things that being can do - count, discern shape and colour, etc.
Personally (and Marnie and I used to have some wild arguments about this) I think that's flawed - in order to get a true value of a creature's intelligence one has to look at how well that creature uses its brainpower within its own environment. Look at dolphin language; for instance - there are two approaches to the research. The first attempts to make use of the dolphins' huge intellect by teaching them to understand human - follow commands; respond to vocal signals, etc. Using this format a researcher can make huge strides in progress. The dolphins want to help; they're having fun, they throw themselves in enthusiastically. This is AFAIK the main means of communication/training for aquarium mammals.
The other way is a lot harder; and shows far less progress. However; I personally feel it will ultimately prove more successful - trying to learn the dolphin's language. Very hard - how do you learn the language of an utterly alien species? Particularly if they're not going out of their way to help.

I think it's an indication of dolphin's intelligence that they're a lot better at learning our language than we with all our technology and education are at learning theirs.

As a different example, horse folk tend follow this second method much more than the first. Yes; training a horse involves essentially teaching it to obey/understand human commands but it goes much farther than that. A good horse trainer understands what the horse is thinking/feeling from one minute to the next; he can communicate with the horse in its language. So do good riders - I'm no expert horseman; but I've never had a problem with 'wild' or 'balky' horses - I know when a horse is tense or fearful; I understand that as herding animals they have a strong social order and can work within that instinctive system.
IOW; horses are incredibly intelligent - at being horses.

OK, this is getting far afield (no pun intended) but I just thought I'd throw out a few more shots about how we look at 'animal intelligence'.

Many folks think that because a species is intelligent enough; it could eventually form a civilization. I submit - based on the above - that a suitably intelligent species, properly equipped for its environment, may have no need for civilization.

Thus; going back to coelurosaurs. I personally believe they were extremely intelligent animals; self aware problem solvers. Quite possibly limited tool users. But they were intelligent at being dinosaurs; which doesn't necessarily include aspects of intelligence we ascribe to humanity.

Whoa - that was probably confusing...too much coffee.

Cheers!