4 ms·
Layman's perspective: we had hints of reasoning from the initial release of ChatGPT when people figured out you could prompt "think step by step" to drastically
by pton_xd 1y ago
Layman's perspective: we had hints of reasoning from the initial release of ChatGPT when people figured out you could prompt "think step by step" to drastically increase problem solving performance. Then yeah a year+ later it was cleverly incorporated into model training.
- impossiblefork 1y agoFine, but to me reasoning is this the where you have <think> tags and use RL to decide what's to be generated in-between them. Of course, people regarded things like GSM8k with trained reasoning traces as reasoning too, but it's pretty obviously not quite the same thing.