This resource is pretty neat and more academic, on mobile you can click into each research paper.
Karl Friston’s nature/natural approach to ai (free energy principle, active inference, spatial web hsml/hstp) may be expanded on more at Davos 2025 next week with expected upcoming atari 10k challenge benchmarks.
Most of his work is academic, some of it with Verses Ai lab. The academic paperworks on Bayesian models is often too mathematical for me but fascinating.
More research paper links:
https://www.fil.ion.ucl.ac.uk/~karl/
https://arxiv.org/search/?query=Karl+friston&searchtype=author&source=header
https://scholar.google.cl/citations?user=q_4u0aoAAAAJ&hl=en
https://www.nature.com/articles/nrn2787
https://arxiv.org/html/2410.10653v1
https://www.aimodels.fyi/papers/arxiv/from-pixels-to-planning-scale-free-active
https://www.mdpi.com/1099-4300/24/3/361
[link] [comments]