An analysis of DeepSeek's R1-Zero and R1

Re:An analysis of DeepSeek's R1-Zero and R1 by ClumsyPilot on Wednesday January 29, 2025 @01:49PM (#42869346)

> The most promising idea is to use reasoning models to generate data, and then train our non-reasoning models with the reasoning-embedded data.

Why is it promising, aren’t you potentially amplifying AI biases and errors?

[ Parent | on HN ]

Re:An analysis of DeepSeek's R1-Zero and R1 by pas on Wednesday January 29, 2025 @08:47PM (#42873822)

it seems to work and seems very scalable, "reasoning" helps to counter biases (answers become longer, ie. the system uses more tokens which means more time to answer a question -- likely longer answers allow better differentiation of answers from each other in the "answer space")

https://newsletter.languagemodels.co/i/155812052/large-scale...

also from the posted article

"""

The R1-Zero training process is capable of creating its own internal domain specific language (“DSL”) in token space via RL optimization.

This makes intuitive sense, as language itself is effectively a reasoning DSL.

"""

[ Parent | on HN ]