Big theropods have "croc"-like brains in so far as they retain the general primitive shape of archosaurs, rather than the highly derived shape of birds. However, all theropods studied so far have a bigger brain cavity than would be expected for a croc of th same body size (that is, a higher EQ (encephalization quotient) than crocs). Non-avian Maniraptorans and ornithomimosaurs have far more bird-like brains (in terms of the shape of the brain cavities) than more primitive forms, and these forms (and tyrannosaurids, actually) seem to have higher EQs than for more basal taxa.
What does that mean for intelligence? Well, if you ask 3 ethologists (animal behaviorists) to define "intelligence" you'll likely get seven different answers... This is because "intelligence" is not one faculty, but is instead our name for a collection of different types of brain operations (memory storage, memory retrieval, spatial relations, problem solving, etc., etc., etc.). It is difficult to measure in humans; it is exceedingly difficult to study in living non-humans. Damned if I can think of an effective testable way of examining it in fossil taxa, beyond the use of proxies like brain size and shape.
So, in any case, dromaeosaur brains were more birdlike than tyrannosaurid or allosaurid brains, but (like those of Archaeopteryx), not as well-developed as those in modern birds.
However, the folks at Jurassic Park have never felt constrained by data when reconstructing the appearance, life habits, and abilities of their dinosaurs. Back when JP existed only in print the raptors were already endowed with cheetah speed and trans-chimpanzee intelligence: not quite faster than a speeding bullet or more powerful than a locomotive, but more Kryptonian than Terran...
Thomas R. Holtz,