"Despite using zero human-curated data, AZR achieves state-of-the-art results on diverse coding and math reasoning benchmarks, even outperforming models trained on large in-domain datasets. This demonstrates the potential for sophisticated reasoning skills to emerge purely through self-play without domain-specific supervision."
wiz21c•11h ago
> "sophisticated reasoning skills"
Does it mean that it uses the data it has to the maximum possible level to produce new reasoning (that add to those produced by less algorithms). IOW, are we still in the realm of: with a given data set, A.I. can produce up to N reasoning capabilities and consequently, can't produce more than that ? IOW, reasoning is bound by knowledge ? And therefore, maybe we could just start from a data/knowledge set in which we add some randomness and self play until some form of reasoning emerge ?
a2128•16h ago
To be clear, this is not a model trained on zero data, this is a pretrained model (Qwen 2.5 trained on 18 trillion tokens) finetuned using self-generated data grounded by a Python interpreter
scotty79•14h ago
I think at this point the initial process of exposing the empty model to all the available domain data in bulk is no longer interesting to many people. It's an obvious first step so it's barely mentioned anymore. What's currently worked on is what you do afterwards to get a useful tool in the end.
ethan_smith•11h ago
The breakthrough here is eliminating the need for human-labeled reasoning data while still achieving SOTA results, which has been a major bottleneck in developing reasoning capabilities.
macrolime•16h ago
Pretty sure OpenAI and/or DeepMind have already been doing something very similar for a while already, just without publishing it.
FieryTransition•15h ago
Agreed, it's a pretty obvious solution to the problems once you are immersed in the problem space. I think it's much harder to setup an efficient training pipeline for this which does every single little detail in the pipeline correctly while being efficient.
squillion•15h ago
Warning: abuse of this technique may cause the model to go blind.
ogogmad•13h ago
Is this a joke about wanking?
QuadmasterXLII•14h ago
For everyone who says “modern incentives forbid publishing negative results,” let this stand as a counterexample!
fotcorn•14h ago
Why do you think it's a negative result?
The table on page 9 shows great results.
ogogmad•13h ago
I think it's a pun. AlphaZero? AlphaNegative.
andy_ppp•12h ago
-273°C isn’t it?
Waterluvian•14h ago
Related to this: has anyone seen a model respond with “oh wait I was wrong…” when you follow-up with a “can you explain why this answer is right?”
I still find that my uses of GPT and others still struggle with a sort of tunnel vision.
gitroom•12h ago
sometimes i feel like the whole self-play thing is kinda the obvious path now but still nuts seeing it actually work better than huge data dumps. you ever wonder how much of progress is just crazy good pipelines versus actual breakthroughs?
nullc•3h ago
Be nice to see some of these run on languages the pretrained model is a little less good at than Python and JS.
mentalgear•17h ago
wiz21c•11h ago
Does it mean that it uses the data it has to the maximum possible level to produce new reasoning (that add to those produced by less algorithms). IOW, are we still in the realm of: with a given data set, A.I. can produce up to N reasoning capabilities and consequently, can't produce more than that ? IOW, reasoning is bound by knowledge ? And therefore, maybe we could just start from a data/knowledge set in which we add some randomness and self play until some form of reasoning emerge ?