The ๐ช๐ฏ๐ท๐ข๐ณ๐ช๐ข๐ฏ๐ต ๐ฐ๐ฃ๐ซ๐ฆ๐ค๐ต๐ช๐ท๐ฆ of ๐ฆ๐ฆ๐ methods causes representational collapse in the latent space, which propagates back to the models outputs.
Anti-Collapsing methods like ๐๐ฒ๐๐๐ฃ๐ or ๐ฆ๐๐๐ฅ๐ฒ๐ด try to solve it in different ways.
However, are they enough?
Check out our latest work at @DeepOriginBio.
Our state-of-the-art ๐๐ข๐๐ผ๐ฐ๐ธ (#docking) and ๐๐ข๐ฆ๐ฐ๐ผ๐ฟ๐ฒ methods led multiple hit discoveries in wet-lab experiments for ๐๐๐ณ๐ฏ, ๐๐ฅ๐๐๐ฐ, ๐๐ฎ๐ฐ๐๐ผ๐ฟ ๐ซ๐, and ๐๐๐ญ๐ณ๐ target proteins.
paper: https://t.co/szWCZVNQ7m
@LindseySnell Too many unfriendly statements or words chosen by the official account of Defense Ministry of Azerbaijan. They use names of regions which do not exist. The "conditional border" is also another chance for them for provocations.
@301arm Is it possible that EU mission intentionally does not want to worsen the situation and refuses admit that such a violation occurred? If not, it is not good for us.
@sharakelyan Yes, the industry has already started to dominate. This can cause migration from academia to industry. By giving different standards, we may speed up this process. But from another perspective, maybe we do not need to hinder scientific development, no matter where.
@sharakelyan because their solutions are comprehensive, they have more reliable justifications and have a wide range of applications. It can also push down the competitiveness of the academy.
2/2
@sharakelyan I think having different standards of publication process for the academy and industry potentially can cause a disbalance of their scientific contributions. From the long-term perspective, people can prioritize industrial solutions over academic ones...
1/2