jparkerholder

3 ms·

1. 1.▲
   by jparkerholder 3y ago

   That is correct! The model generates each frame on the fly.
2. 2.▲
   by jparkerholder 3y ago

   If we did a good job then the paper should be written in a way that is digestible. When you don't understand things, follow the references to learn more (and there's probably videos covering most of the components we use). In the
3. 3.▲
   by jparkerholder 3y ago

   The bigger issue is lack of generating novel content rather than a total "unravel". We focus on OOD images because our motivation is generating diverse environments, but these are much harder to play for longer vs images closer to
4. 4.▲
   by jparkerholder 3y ago

   Hey author here :) First, tough crowd, love it, always great to get feedback because we are actively working on improving the model. We are very happy to admit it is not perfect, but given not many people thought this was possible a year ag
5. 5.▲
   by jparkerholder 3y ago

   Author here :) Re: 1) typically no, but of course it can hallucinate just like LLMs. 2) Agreed but the key point missing is Dreamer is trained from an RL environment with action labels. Genie is trained exclusively from videos and learns an