Leshem Choshen ๐ค๐ค
@lchoshen
๐ฅ Collaborative LLMs (co-created model merging)
๐ฅ Opinionatedly sharing #ML & #NLP
๐ฅ We owe science an alternative hype
@IBMResearch & @MIT_CSAIL
ID: 1006797311593377792
https://ktilana.wixsite.com/leshem-choshen 13-06-2018 07:15:59
7,7K Tweet
3,3K Takipรงi
571 Takip Edilen
Human feedback is a key component for improving the capabilities and safety of AI models but itโs currently concentrated in a few companies. We want to change this, so weโve outlined a vision for creating an ecosystem for sharing and collaborating on open human feedback data๐๐ผ
This is the one paper to read if you care about human input into AI systems. You know who else agrees we have to collect and leverage human feedback in open source AI? That's right Yann LeCun and definitely Soumith Chintala
One thing I've learnt from Ramon Astudillo model releases RLHF multiple times! We can't even train once on open data...
One thing I didn't really think about before this, when I talk to a model and it generates an image, the image is mine. Why isn't it the case when it is text? Well, isn't it? t\h Ramon Astudillo Yangsibo Huang others