you could test the temps with an external temperature probe...
or, record what each temp says at idle, then start up prime 95 or something. Then record what they say for load. One should have a big difference, and one might go up maybe 1 or two degrees. The one with the larger difference should be the correct one