Understanding R1-Zero-Like Training: A Critical Perspective
1–10 of 23 posts
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#2If the base models already have the “reasoning” capability, as they claim, then it’s not surprising that they were able to get to SOTA using a relatively negligible amount of compute for RL fine-tuning.
I love this sort of “anti-hype” research. We need more of it.
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#3So they achived R1-Zero like performance without those long CoT that sometimes never end/are impacting inference time with fraction of fine tunining resources?
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#4[flagged]
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#5So they achived R1-Zero like performance without those long CoT that sometimes never end/are impacting inference time with fraction of fine tunining resources?
No, they still have "", but it's shorter by removing part of a term.
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#6I've seen the same "Superficial Self-Reflection" mentioned in their linked blog post[0] as well, where the conclusion doesn't naturally follow the output of the thinking tokens. I think people are fooled by this, but if you take the time to inspect the "chain of thought" tokens they often don't match the final output answer.
I don't deny that performance for certain logic tasks goes up with these models but I don't fully understand what role the thinking tokens take in these cases.
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#7So they achived R1-Zero like performance without those long CoT that sometimes never end/are impacting inference time with fraction of fine tunining resources?
No, they still have " ", but it's shorter by removing part of a term.
That's what I mean, those CoT are never ending currently until you run out of context.
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#8Overall the industry needs more review, less hype. I was shocked to find out SWE-verified [0] is all but verified.
[0] benchmark used by all major vendors to "showcase" coding ability, turns out to be https://www.youtube.com/watch?v=QnOc_kKKuac
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#9Earlier quoted context omitted.
No, they still have " ", but it's shorter by removing part of a term.
That's what I mean, those CoT are never ending currently until you run out of context.
I'm not sure if you're talking conversationally and I'm taking it as a technical query, or you're saying CoT never terminate for you and asking for input, or asking what the paper implies about CoT, or relaying that you understand the papers claim that this method net reduces CoT length.
Re: Understanding R1-Zero-Like Training: A Critical Perspective
#10I've seen the same "Superficial Self-Reflection" mentioned in their linked blog post[0] as well, where the conclusion doesn't naturally follow the output of the thinking tokens. I think people are fooled by this, but if you take the time to inspect the "chain of thought" tokens they often don't match the final output answer. I don't deny that performance for certain logic tasks goes up with these models but I don't f…
I heard that even just getting the model to print a bunch of whitespace ("think for longer") improves the quality of the final response, because some kind of processing is still happening internally?