> Neural networks can learn algorithms, not just patterns. The 777-parameter model didn’t memorize. It couldn’t. It discovered a general procedure for addition that works on numbers it had never seen.
If that's true, I would have expected accuracy to be 100%. Why isn't it 100% accurate?
Executing an algorithm that you've learned and that is true should have 100% accuracy. Otherwise, I think you either haven't actually learned the algorithm or you learned an incorrect algorithm.
Ah, I see. As for my personal neural network: I learned long division in school and got the best possible grade in maths, and I still don't think I had 99.63% accuracy on long division.
7 comments
[ 0.29 ms ] story [ 29.7 ms ] thread> Neural networks can learn algorithms, not just patterns. The 777-parameter model didn’t memorize. It couldn’t. It discovered a general procedure for addition that works on numbers it had never seen.
If that's true, I would have expected accuracy to be 100%. Why isn't it 100% accurate?