For any researcher in life sciences having a hard time staying current in their field: Paperzilla now has the PubMed literature index as a source.
โ bioRxiv
โ medRxiv
โ ChemRxiv
โ PubMed
I'm using Codex/ChatGPT myself (a lot), and I'm now using our own Paperzilla plugin to chat every morning about the new papers I am interested in. The only way to stay current and stay sane :)
I love it when a paper comes with open-source code. And in this case, even a website!
You might know the image segmentation models we had back in the day. Photo in, and a pixelated image with boundaries out.
This recipe takes that to a whole new level by using an image generator to analyze an image.
It uses an image generator to analyze an existing image by having it paint the answer. For example, it can paint all car pixels red and all road pixels blue. That color map is then decoded into object labels.
Paper and summary ๐๐
I love it when a paper comes with open-source code. And in this case, even a website!
You might know the image segmentation models we had back in the day. Photo in, and a pixelated image with boundaries out.
This recipe takes that to a whole new level by using an image generator to analyze an image.
It uses an image generator to analyze an existing image by having it paint the answer. For example, it can paint all car pixels red and all road pixels blue. That color map is then decoded into object labels.
Paper and summary ๐๐
This fascinating paper showed up in my Paperzilla morning research brief (inside Claude). Easy-to-read summary and links to paper and repo:
https://t.co/Fd6EHYTuj5
@Hetzner_Online@levelsio@digitalocean Maybe let existing customers scale their VPSs first. I assume you have growth issues, but letting existing customers suffer from that is really bad business practice.
What happens if we add some quantum hardware to an LLM? Will it be faster, better, more efficient?
That is what the researchers in this study asked themselves. The result? It removes 16.33% of the modelโs parameters, with roughly comparable next-token prediction performance.
I don't know enough about this to say whether we should celebrate it. But if you are into this, make sure to check out the paper ๐