Newly published paper on the Forward-Forward algorithm (FF), a technique for training neural networks that uses two forward passes of data through the network, instead of back-propagation, to update the model weights. https://t.co/ghq7ud9HgA
new proof, that if correct, comes close to proving the nonexistence of Landau-Siegel zeros (progress towards the Generalized Riemann Hypothesis) https://t.co/tapXfsMaGO
AI is moving fast; a couple of months ago from text-to-image with stable diffusion and Dall-E, now with text-to-video (Make-A-Video) https://t.co/i0ZWPuW1Vv
a colorful (philosophical) paper describing 'data alchemy', a manifesto with ten propositions regarding diffusion as a narrative for discussing data https://t.co/j69vowzLg5