Ask HN: How do you catch up to the research of LLMs/Transformers etc.?
I have been trying to catch up on research that has happened in the last 5 to 10 years on transformers, GANs, LLMs, etc. There are some amazing resources out there explaining some of the aspects. I also started reading original papers for better context. However, most of them refer to previous papers and I feel like I am falling into the rabbit hole and discovering more research.
How do you catch up to the fast-moving research?
Here are some resources I use:
1. https://paperswithcode.com/
2. Two minutes papers ( Youtube )
3. https://github.com/diff-usion/Awesome-Diffusion-Models
4. https://www.litmaps.com/ ( for finding references )
5 comments
[ 2.2 ms ] story [ 22.7 ms ] threadhttps://karpathy.ai/zero-to-hero.html
Mastodon: @arxiv_ai@creative.ai
https://metacurate.io/
Research tools: https://tools.kausalflow.com/tools/
https://news.ycombinator.com/item?id=38652736
- follow (amongst hundreds/thousands of other accounts..): ak, aran komatsuzaki, lucidrains, rivershavewings, wizardlm, blinkdl (rwkv), greg gerganov, thebloke, yannic kilcher, yann lecun, hazyresearch, igor carron, calccon/charles martin, birchlabs, jeremy howard, geohot, patrick kidger, gwern, adam nemecek