On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models · HackerPapers discussions

On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models — HN submission

Thu, 30 Nov 2023 19:07:53 GMT

Algorithmic Information Theory for RL Controllers and RNN World Models

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models — mentioned by gwern

Wed, 16 Aug 2023 17:36:21 GMT

This is an old idea dating back to the earliest algorithmic information theory work in the '50s/60s like Ray Solomonoff on universal induction: given a powerful learning algorithm, create AI by encoding all possible data as a binary stream (with no further preprocessing or engineering), and simply predict each successive bit. This leads to AIXI, the compression-as-intelligence paradigm (and online learning / meta-learning), and most recently, the success of self-supervised learning & generative modeling. See for example Burfoot https://arxiv.org/abs/1104.5466 or Schmidhuber…

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

On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models — mentioned by hardmaru

Wed, 18 Apr 2018 07:32:16 GMT

> How did you get into contact with Schmidhuber for co-authoring? What stage was the research at when he joined? The first time I discussed this topic with Jürgen Schmidhuber was at NIPS 2016, when he gave a talk about "Learning to Think" [1], during the break at one of the sessions, and we kept in contact afterwards. > Were you expecting the net to generalize from dream to reality, before you wrote the paper, or did this materialize during experimentation? When I tried this, I didn't expect this to work at all, to be honest! And in fact, as discussed in the paper, it didn't work at the…

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

On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models — mentioned by nrmn

Wed, 23 Dec 2015 19:44:54 GMT

I've been trying to read a paper a day since midsummer. These are a few of the interesting, for me personally, since then: Generating Sequences With Recurrent Neural Networks - http://arxiv.org/abs/1308.0850 Older one, but important to understand deeply since other recent ideas have come from this! Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks - http://arxiv.org/abs/1511.06434 Unitary Evolution Recurrent Neural Networks - http://arxiv.org/abs/1511.06464 State of the Art Control of Atari Games Using Shallow Reinforcement Learning…

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

On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models — HN submission

Tue, 01 Dec 2015 13:10:01 GMT

On Learning to Think: Algorithmic Information Theory

155 points; 3 discussion replies on HN. Paper and source history. A submission is not an endorsement.