Show HN: I RL-trained an agent that trains models with RL (for –$1.3k)

Re:Show HN: I RL-trained an agent that trains models with RL... by lumost on Tuesday July 14, 2026 @09:56AM (#48906969)

I think the counter point for these projects is that you may not need a deep understanding if you can measure the outcome. While this may not be true every time today, it plausibly will be in the future - making the activity worthwhile.

[ Parent | on HN ]

Re:Show HN: I RL-trained an agent that trains models with RL... by Danau5tin on Tuesday July 14, 2026 @10:22AM (#48907376)

Yes I do agree with this. I believe we are shifting from "make the model good" (prompt/context engineering, etc) to "define good for the model" (success criteria/rubrics). Over time I believe this will become increasingly obvious (as long as model capabilities continue to increase).

[ Parent | on HN ]

Re:Show HN: I RL-trained an agent that trains models with RL... by saberience on Tuesday July 14, 2026 @09:58AM (#48906991)

Well, you say that, but when "measuring" anything in RL, that measurement itself is not always obvious.

That is, creating the scoring system/judge models etc for RL is not easy at all. You can easily create an RL loop which is getting better and improving its scores, but actually the result is totally garbage, because you're measuring the wrong thing.

[ Parent | on HN ]

Re:Show HN: I RL-trained an agent that trains models with RL... by babelfish on Tuesday July 14, 2026 @01:03PM (#48909868)

What do you see wrong with the OPs reward design?

[ Parent | on HN ]

Re:Show HN: I RL-trained an agent that trains models with RL... by willleeney on Monday July 20, 2026 @07:22AM (#48977254)

They said that the we are moving towards " `define good for the model` (success criteria/rubrics)" and you've just gone no no, you don't understand you need to measure the right thing. Feels like you've just completely agreed with them?

[ Parent | on HN ]