analiza aiqbench
This research introduces a neural Langevin machine that utilizes local, asymmetric learning rules to achieve generative capabilities with biological relevance. The model demonstrates a distinct memorization-to-generalization transition and offers practical utility in continuous phase space exploration and image denoising. Its reliance on local neural signals rather than global backpropagation suggests potential for more efficient, neuro-inspired architectures.
Neural Langevin Machine: a local asymmetric learning rule can be creative
arXiv:2506.23546v2 Announce Type: replace-cross Abstract: Fixed points of recurrent neural networks can be leveraged to store and generate information. These fixed points can be captured by the Boltzmann-Gibbs measure, which leads to neural Langevin dynamics that can be used to find them for generative learning of a real dataset. We call this type of generative model a neural Langevin machine, which derives an asymmetric and firing-rate-speed adjusted learning rule requiring only local neural signals, thereby bearing biological relevance in terms of local predictive learning. An interesting out-of-equilibrium regime of the generative process is revealed, together with a memorization-to-generalization transition with increasing training data size. The neuro-inspired machine can also realize a continuous exploration of the phase space for different kinds of generative images and can denoise a corrupted image as well.
Abstract:Fixed points of recurrent neural networks can be leveraged to store and generate information. These fixed points can be captured by the Boltzmann-Gibbs measure, which leads to neural Langevin dynamics that can be used to find them for generative learning of a real dataset. We call this type of generative model a neural Langevin machine, which derives an asymmetric and firing-rate-speed adjusted learning rule requiring only local neural signals, thereby bearing biological relevance in terms of local predictive learning. An interesting out-of-equilibrium regime of the generative process is revealed, together with a memorization-to-generalization transition with increasing training data size. The neuro-inspired machine can also realize a continuous exploration of the phase space for different kinds of generative images and can denoise a corrupted image as well.
Neurons and Cognition (q-bio.NC); Disordered Systems and Neural Networks (cond-mat.dis-nn); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
arXiv:2506.23546 [q-bio.NC]
(or arXiv:2506.23546v2 [q-bio.NC] for this version)
Text integral sub licența deschisă a publicației · 152 cuvinte
Subiecte conexe
Neural Langevin Machine: a local asymmetric learning rule can be creative
arXiv · 18 sept. 2026
Humanity’s Last Invention — Richard Socher of Recursive
Latent Space · 14 sept. 2026
At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI
NVIDIA · 20 iul. 2026
AI Futures: Racing to Lose, A sermon
LessWrong · 28 sept. 2026