logoalt Hacker News

jononortoday at 6:39 PM0 repliesview on HN

If you have good feedback signals, like tests/benchmarks/etc, then it is potentially better to do multiple turns where model uses that to adjust code. Which might not need as smart a model.