jgrizou
Joined 9 years ago
Karma
15
Submissions
27
Member since
Jan 5
https://jgrizou.com/
View on Hacker NewsJoined 9 years ago
It feels like this proposal maintain the paper as the central artefact, but maybe the paper should be the side/byproduct. In many computer science fields it is clear that reproducible code should come first for example.
A in-model memory approach is probably still deep research but maybe a Rag-like pipeline could work in some instances
On waking up, we show you the video of what happened. You feel the same person as yesterday when you went to bed. Are you?