3 ms·
nice! Training models using reward signals for code correctness is obviously very common; I'm very curious to see how good things can get using a reward signal
by justusm 1y ago
nice! Training models using reward signals for code correctness is obviously very common; I'm very curious to see how good things can get using a reward signal obtained from visual feedback
- grace77 1y agoAs are we, seems like the natural next step