Reinforcement Learning (RL) is a powerful method for controlling dynamic systems, but its learning mechanism can lead to unpredictable actions that undermine the safety of critical systems. Here, we propose RL with Adaptive Regularization (RL-AR), an algorithm that enables safe RL exploration by combining the RL policy with a policy regularizer that hard-codes the safety constraints. RL-AR performs policy combination via a "focus module," which determines the appropriate combination depending on the state--relying more on the safe policy regularizer for less-exploited states while allowing unbiased convergence for well-exploited states. In a series of critical control applications, we demonstrate that RL-AR not only ensures safety during training but also achieves a return competitive with the standards of model-free RL that disregards safety.

We show that these methods fail catastrophically when limited to small replay buffers or during incremental learning, where updates only use the most recent sample without batch updates or a replay buffer.

Deep models have demonstrated remarkable performance in time series forecasting.

In this paper, we provide sufficient conditions of DMM's stochastic stability as defined in the context of dynamical systems

The EU's Global Gateway plan is challenging China's Belt and Road Initiative to influence Africa, by providing funding that will expand access to electricity. Nearly 600 million Africans--half the continent's population--are without electricity, largely because of the continent's limited distribution network, and Africans make up the vast majority of those worldwide without electricity access. But the European Union wants to change this. At the end of September, the president of the European Commission, Ursula von der Leyen, announced a €545 million ($636 million) investment package to support renewable energy and electrification in Africa. New EU-funded projects will include a high-voltage transmission line in Côte d'Ivoire, the electrification of hundreds of rural communities in Cameroon, the exploitation of wind and hydro energy in Lesotho, and the installation of mini-grids in remote areas of Madagascar.

We consider the problem of estimating states ( e.g., position and velocity) and physical parameters ( e.g., friction, elasticity) from a sequence of observations

How Hong Kong gave rise to Labubu and a designer toy movement now shaping global culture. The following sentence might make a globalist cry out for joy: A toy that is manufactured by a Chinese company in Vietnamese factories, designed by a Dutch artist in Belgium, inspired by indie toy culture in Hong Kong, and made viral thanks to a Thai K-pop star, has turned into the biggest Gen-Z cultural trend of 2025. That abomination of a sentence is the story of Labubu, the creepy-cute stuffed monster that swept the world this summer. You must have seen the trend by now, but most people are still unaware of the global, decade-long story that led up to it. Last week, I published a feature story about my journey into the heart of Labubu, how this cultural mania moment was created, and where it may go from here.

We propose that these limitations arise because MPNNs only pass two-body messages leading to a direct relationship between the number of layers and the expressivity of the network.