blitz retweeted
New blog!
This one covers my experiments on fine-grained credit assignment with an agentic judge, and how it affects the models' performance and reasoning quality and some thoughts on RL science.
Link below, please do check it out, and RT/QT if you like it:)
my blue check is about to expire, so figured id throw this out here as one last Hail Mary.
looking for ai engineer / founding engineer roles.
blr (india) or remote.
python, typescript, llms, backend, agent infra.
DM if hiring.
nitter.cf/messages/compose?recip…
trying codex with unity. and watching it struggle to setup animation properly for humanoid model is so funny
5
86
hi we are hiring applied ml intern/s
requirements
- strong cs + ml fundamentals
- prior experience building ai agents or harnesses that actually work
- high agency and first principles thinking obv
- bonus points if youve worked on: rl envs, eval frameworks or verifiable systems with lean
to apply just dm me or reply with your work, projects(github links too) and resume(optional)
in person preferred (in blr) but can consider remote if youre v v good
thanks and rt for good karma :D
i didn't have pc back then so i was playing most of the games from the list on mobile, now that i have it i gotta play more games
1
7
272














