Nick

just emergent behavior

[Nick @nickwal]

@RhysSullivan Depends on the model usually no but in our benchmarks there are models where it helps a ton but they usually have weaker reasoning capabilities

[Nick @nickwal]

@benhylak oh yes!!!

[Nick @nickwal]

@leedsharkey @GoodfireAI this is so cool! great work

[Nick @nickwal]

Oh my god the whale is back

[Harry Partridge @part_harry_]

Human-in-the-loop RL is necessarily done at group size 1; you cannot do a group of rollouts with only one human. i.e. there is no baseline for you to subtract for each input prompt. This is by far the most interesting and under-discussed part of this announcement.

The same was true for their tab-completions model. From the wording in their posts, it sounds like they are using plain REINFORCE (no mention of value functions) with a large batch size + re-evaluating each checkpoint to guard against high variance. Cursor is implicitly revealing an important empirical result: with a large enough batch size, simple REINFORCE just works, no baseline needed. In other words, large scale continual learning is solved.

[Nick @nickwal]

@leonardtang_ sick

[Bartosz Naskręcki @nasqret]

It finally happened-my personal move 37 or more. I am deeply impressed. The solution is very nice, clean, and feels almost human. While testing new models in the last few weeks, I felt this coming, but it's an eerie feeling to see an algorithm solve a task one has curated for about 20 years. But at least I have gained a tool that understands my idea on par with the top experts in the field. And I am now working on a completely new level. My singularity has just happened… and there is life on the other side, off to infinity!

[Nick @nickwal]

@xeophon @PrimeIntellect congrats!

[Nick @nickwal]

@si_pbc sick

[Nick @nickwal]

@jxnlco @OpenAI @romainhuet @OpenAIDevs let’s go!!! congrats

[Nick @nickwal]

@ProximalHQ Congrats!

[David @DavidSHolz]

5 million humanoid robots working 24/7 can build Manhattan in ~6 months. now just imagine what the world looks like when we have 10 billion of them by 2045. now imagine the year 2100.

[Nick @nickwal]

@GoodfireAI this is so cool

[Nick @nickwal]

very inspiring vision for the future of research

the hosted training has been incredible to iterate with

[Prime Intellect @PrimeIntellect]

Introducing Lab: A full-stack platform for training your own agentic models

Build, evaluate and train on your own environments at scale without managing the underlying infrastructure.

Giving everyone their own frontier AI lab.

[Prime Intellect @PrimeIntellect]

Introducing Lab: A full-stack platform for training your own agentic models

Build, evaluate and train on your own environments at scale without managing the underlying infrastructure.

Giving everyone their own frontier AI lab.

[Nick @nickwal]

@thdxr providing*

[Nick @nickwal]

@thdxr I believe this is regarding intra-turn prefill where you precondition the response by proving the first few tokens for that turn of the assistant response

I believe this is unrelated to prior turns in the chat format

[Nick @nickwal]

@latkins yooooo amazing work!

[Nick @nickwal]

@msfeldstein fetch tool / web search is very brittle (constantly hangs for me) and could use some improvements regarding what it’s actually searching