Jin: Policy gradient algorithms function as weighted supervised fine-tuning
Roger Jin · What is an RL environment? w/ Nous Research's Roger Jin · Apr 29, 2025 · at 5:59
This episode carries Roger Jin's own address, with nobody on the show putting questions to them. It still counts as said, and it is kept out of every score on their page.
Roger Jin of Nous Research connects policy gradient reinforcement learning algorithms such as REINFORCE and GRPO directly to supervised fine-tuning.
0:00 / 0:18exact quote · 18.5s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“If you kind of, like, look at, if you kind of stare at, like, this part it sort of looks like just, like, weighted supervised fine-tuning, right? Like, you have this, like, log of, like, the probability of a token and, like, some, like, weight on it and reinforces, like, the return, but it doesn't have to be that, like, you can replace it with, like, some advantage like, GRPO does that, right?”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Roger Jin
Jin: RL enables models to surpass expert labelers and develop self-direction
“The model outperforming expert labelers is, is possible. The model learning, like, self-direction is, like, expected. And yeah, we've seen, like, kind of cool emergent behaviors with, like, you know, like, O-one, O-three, R-one, kind of, like, these, like, thi…”
Jin: Supervised Learning Cannot Train Models by Demonstrating Bad Examples
“There's, like, a kind of deficit, which is, like, there's no way to demonstrate what is bad in supervised learning. You kind of just, you have to tell the model very well, like, what is good.”
Jin: Supervised Models Rarely Outperform Expert Human Labelers
“There are, like, kind of surprising results, like weak to strong generalization, but, like, in general, like, the model outperforming the expert labelers would be, like, kind of very surprising.”
Jin: Open source needs a standard to scale RL environments
“And so, like, there just needs to be some kind of, like, standard for open source developers to all, like, work together to, like, build up this, you know, to scale environments up to, like, millions and millions of environments.”
Jin: Fusing inference and scoring natively enables multi-step and multi-agent RL
“Collect trajectories is a fusion of both these. It handles both inference and scoring, and we deliberately chose that because, like, what happens when you try to, like, do, like, multi-turn, or, like, multi-agent with, like, a separate score function? Then thi…”
Jin: Language models map directly to reinforcement learning policies
“So the states are, like, the text prefixes, so the initial states, like, the prompt the actions are the next tokens that means, like, a language model is, like, exactly what a policy is, right? A policy maps a state to a probability distribution of our next ac…”