It's not a weighted average though. I see this repeated all the time. Between synthetic data, custom produced training data by experts, post-training regimes etc the models are far more diverged from that baseline at this point, at this point it's mostly a right shifted curve.
replyI hesitated to write that line because I don't know the math. So thanks for clarifying. Main point is the output is along some distribution and I'd say AI-evangelist themselves see "good enough" across everything and anything as a feature.
reply[dead]
reply