Artur d'Avila Garcez: First-order reasoning and planning t...| Robocrunch AI

First-order reasoning and planning together with fast learning from low-level data requires: "an architecturally significant weaving of neural and symbolic parts, neither of which would be sufficient to solve the problem" @Grady_Booch at least for now.
— Artur d'Avila Garcez (@AvilaGarcez) May 9, 2022

Related

https://arxiv.org/abs/2106.04537

My lab has been focusing on neural networks that learn planning algorithms end-to-end. Neural nets can solve reasoning and planning problems (e. ...

Shared by Tom Goldstein at 3/9/2022 0-9

https://nvda.ws/3Etqa75

Combining human intuition and ML techniques can help identify optimal neural network architectures for your problem: https://t.co/BZ7th7oqoR

Shared by NVIDIA AI at 4/18/2022 2-8

http://arxiv.org/abs/2112.00114

Another solution is to use in-context learning to induce rational generation [Nye et al. https://t.co/87sE9kiGI4, Wei et al. https://t.co/q70C9X ...

Shared by Yuhuai (Tony) Wu at 3/29/2022 0-3

As it turned out, solving the problem of learning hierarchical representations and complex functional dependencies was a much more important iss ...

Shared by Yann LeCun at 5/14/2022 0-8

https://sites.google.com/view/nesy-2022/home

16th International Workshop on Neural-Symbolic Learning and Reasoning (NeSy2022). Call for papers: https://t.co/Gnj8xbQseg

Shared by Artur d'Avila Garcez at 2/14/2022 2-3

"chain of thought prompting for neural network" #dalle

Shared by Aravind Srinivas at 4/12/2022 1-5

768