Repenomamus giganticus
I'm sure everyone has already heard about the discovery of a dinosaur-eating mammal from the Cretaceous. Here are some photos of the creature from the announcement in Nature.
I'm also sure that the scientifically-literate readers of the Panda's Thumb and Pharyngula won't be terribly surprised by this—the mammal-like reptiles are older than the dinosaurs, and mammals were contemporaries of the dinosaurs. The surprise is that what was discovered was a bandy-legged brute that was relatively large (about the size of your average dog) and was so danged uppity that it had been eating dinosaurs. That's a bit more temerity than had been expected from our long-lost relatives.
Here's the skull of the Cretaceous carnivore Repenomamus giganticus…good strong jaws there.
And here's the whole skeleton:
Repenomamus is the genus, and there are two known species. The big guy above is Repenomamus giganticus, and the other, Repenomamus robustus, is smaller, about the size of an opossum. It's the smaller individual that was found to contain a Psittacosaurus skeleton. It's not certain how the animal acquired this meal, but the authors come down on the side of predation.
It is not easy to assess whether Repenomamus was a predator or scavenger. Scavengers are relatively rare among mammals—among extant carnivorous mammals, only two species of hyenas are habitual scavengers. Compared to their hunting cousins, these hyenas have smaller second upper incisors and less jaw muscle leverage, which probably reflect their inability to capture and handle live prey. In contrast, the enlarged incisors and strong jaw muscles of Repenomamus are well shaped for catching prey, favouring it as a predator rather than a scavenger.

The postcranial skeleton of R. robustus (IVPP V13605). a, Ventral view of the skeleton and its stomach content. Associated partial skull and lower jaws not illustrated. b, Buccal view of lower teeth of the juvenile Psittacosaurus. c, Lingual view of two lower teeth of the juvenile Psittacosaurus. d, e, Close-up view of the stomach content (d) with identified elements highlighted in colour (e). Ca, caudal vertebra; Cl, clavicle; Fe, femur; Fi, fibula; Hu, humerus; Il, ilium; Is, ischium; L1, first lumbar vertebra; Ma, manubrium; Mu, manus; Pe, pes; Ra, radius; St, sternum; T20, twentieth thoracic vertebra; Ti, tibia; Tr1, Tr10, Tr14, first, tenth and fourteenth thoracic ribs; Ul, ulna. Measurements of the juvenile Psittacosaurus (length in mm): humerus, 21; radius, 18; ulna, 19; tibia, 36; and fibula, 35.
As was pointed out in the comments, here is the obligatory reconstruction restoration:

Hu Y, Meng J, Wang Y, Li C (2005) Large Mesozoic mammals fed on young dinosaurs. Nature 433:149-152.






PZ,
Seeing that several cynodonts were carnivorous and probably active enough to hunt young dinosaurs (eg Cynognathus or probably even Thrinaxodon) is this really all that surprising?
Even if the above answer is no, you must admit that this is way cool!!