Post

we're approaching superhuman coding and math but somehow LLMs are still mid at "writing"? this is undertheorized

@codehappy.bsky.social

it is easy to directly do RL with verifiable reward in those domains; there is a correct answer and model responses can be scored automatically. it is also easy to synthesize new problems in those domains, and to estimate difficulty of each problem writing is different. reward function is messy

13/5/26, 3:40 am0 reposts6 likes

brennan

@brennan.computer

13 May

+1 over the years I've 'written' a lot of stories and tried to build tooling for recursively capturing style and such, and it's just so hard to carrot (or stick) behavior towards such a squishy definition of a goal