6 pointsby tbharath6 hours ago2 comments

OsrsNeedsf2P6 hours ago
Yes, that's why good datasets are important. But usually the difference between model generations is architectural improvements
- verdverm6 hours ago
  late training is where a lot of capability gains come from too
  - tbharath4 hours ago
    Interesting, how does that work?
    verdverm4 hours ago
    the art of reinforcement learning
verdverm6 hours ago
How much do we assume human written code for training the original models is free of human made slop? We all know we all take shortcuts and have code we are not proud of. If they are I deed averaging machines, is it possible their output is the average of human output?
I'd argue that the way attention works plays into the slop too
- bdangubic4 hours ago
  30 years of looking at human written code, not great :) in the early years for me (late 90’s) code was just solid. dotcom boom happened and “overnight” everyone was a fucking coder. there are too many of us and average coder is so bad that you would never hire them. not only that, even at 90 percentile is fairly terrible coder. AI has absolutely been training on this garbage human code so it was terrible at coding initially… but it is evolving now on its own