luiscosio

4 ms·

1. 1.▲
   by luiscosio 9mo ago

   Thanks for the insight! I think this is exactly why some of the work is moving away from “assume unrestricted API inference forever.” For example, we’re prototyping ideas like air-gapped or very low-bandwidth inference gateways, where inter
2. 2.▲
   by luiscosio 9mo ago

   I’m working on Security Level 5 (SL5), which is basically “nation state grade security” for frontier AI systems. The core idea is that if a model’s weights or training artifacts could enable catastrophic harm, you should treat them like top
3. 3.▲
   Llama 3.1(llama.meta.com)

   437 points by luiscosio 2y ago|269 comments
4. 4.▲
   Llama3(1) 405B (basemodel) has leaked already in 4chan(twitter.com)

   3 points by luiscosio 2y ago|0 comments
5. 5.▲
   by luiscosio 7y ago

   If someone needs help with cloud computing or ML, let me know!
6. 6.▲
   by luiscosio 7y ago

   There is also: https://github.com/deepfakes/faceswap With a big development community and interesting results.
7. 7.▲
   by luiscosio 10y ago

   There is something similar for Tensorflow: https://github.com/calvinschmdt/EasyTensorflow "This package provides users with methods for the automated building, training, and testing of complex neural networks usin
8. 8.▲
   Patent for[..]in-line viewing of files over a network such as the Internet(google.com)

   1 points by luiscosio 16y ago|0 comments
9. 9.▲
   by luiscosio 16y ago

   Vladimir, if you are interested in working on a startup in México as paid intern. Ping me.
10. 10.▲
    by luiscosio 16y ago

    Pretty serious considering a large amount of users use the same password among several sites: http://www.securityweek.com/study-reveals-75-percent-individ...