Because we do not have time to read everything.
Daily Arxiv summaries: https://yobibyte.github.io/arxiv_compressed.html
Compressor is an LLM-based scientific literature / talks summarisation project started by yobibyte. It is heavily relying on llama.cpp and HuggingFace models.
Compressor is under active development, you are entering unchartered waters when using it.
I will be happy to any feedback / feature requests, and, please, send PRs.
-
Get arxiv link, summarise.
-
Get all papers submitted to Arxiv at a date (usually published today). Summarise each.
-
Get a pdf, summarise. Not yet implemented.
-
Get an audio of a talk, get a script, summarise. WIP.
-
Summarise all papers accepted to some conference on OpenReview.
-
Summarise all talks of a particular conference. Future plans.
Crawler -> Compressor -> Reporter
- Current version does summarisation based on abstracts only. Add full-text support.
- Better exception handling. Right now, postprocessing LLM outputs might fail from time to time requiring rerunning the compressor.