Human-Timescale Adaptation in an Open-Ended Task Space · HackerPapers discussions

Human-Timescale Adaptation in an Open-Ended Task Space — mentioned by amf12

Sat, 11 Feb 2023 17:53:57 GMT

These are based on published research by Deepmind, which is an AI research lab. They don't build products to "demo" anything to the public. They research and publish peer-reviewed papers. The only thing they can "demo" is the simulation of the model, which they have published as videos on YouTube. Here's the paper if you're curious btw: https://arxiv.org/pdf/2301.07608.pdf

Paper and discussion history. A mention is not an endorsement.

Human-Timescale Adaptation in an Open-Ended Task Space — mentioned by margorczynski

Sat, 11 Feb 2023 14:50:05 GMT

It's behind a paywall, any way to go past it? Also would be great if there was an actual paper or more serious article about it. I see a lot of this type of articles recently and most of them are wrong/hype/etc. and the only valuable part are the links to the sources. EDIT: Found these with Google: https://arxiv.org/abs/2301.07608 https://www.youtube.com/watch?v=U93bUQ1roiw

Paper and discussion history. A mention is not an endorsement.

Human-Timescale Adaptation in an Open-Ended Task Space — mentioned by gardenfelder

Fri, 20 Jan 2023 16:43:08 GMT

paper: Human-Timescale Adaptation in an Open-Ended Task Space https://arxiv.org/abs/2301.07608 Edit: Reddit: https://www.reddit.com/r/MachineLearning/comments/10g281u/r_...

Paper and discussion history. A mention is not an endorsement.