We propose to learn to distinguish reversible from irreversible actions for better informed decision-making in Reinforcement Learning (RL). From theoretical considerations, we show that approximate reversibility can be learned through a simple surrogate task: ranking randomly sampled trajectory events in chronological order. Intuitively, pairs of events that are always observed in the same order are likely to be separated by an irreversible sequence of actions. Conveniently, learning the temporal order of events can be done in a fully self-supervised way, which we use to estimate the reversibility of actions from experience, without any priors.We propose two different strategies that incorporate reversibility in RL agents, one strategy for exploration (RAE) and one strategy for control (RAC). We demonstrate the potential of reversibility-aware agents in several environments, including the challenging Sokoban game. In synthetic tasks, we show that we can learn control policies that never fail and reduce to zero the side-effects of interactions, even without access to the reward function.

Two decades ago, social media promised to connect people with pals far and wide. Twenty years online has left us turning to AI for kinship. IRL companionship is the future. Anyone looking for a vibe check on the populace's current feelings about AI would do well to check out the walls of the New York City subway system. This fall, alongside posters for everything from dating apps to Skechers, a newcomer made its debut: Friend .

Health Psychology Mental Health Go ahead and swear--it's good for your health Cursing can boost your workout, mood, and even confidence. Breakthroughs, discoveries, and DIY tips sent every weekday. Yelling a properly timed swear word isn't only emotionally satisfying--it may have real physical and psychological benefits . In fact, a well-voiced expletive might even help take you to the next level during a particularly strenuous workout. "In many situations, people hold themselves back--consciously or unconsciously--from using their full strength," explained Richard Stephens, a psychologist at Keele University in the United Kingdom.

Continuous attractors offer a unique class of solutions for storing continuous-valued variables in recurrent system states for indefinitely long time intervals.Unfortunately, continuous attractors suffer from severe structural instability in general---they are destroyed by most infinitesimal changes of the dynamical law that defines them.This fragility limits their utility especially in biological systems as their recurrent dynamics are subject to constant perturbations.We observe that the bifurcations from continuous attractors in theoretical neuroscience models display various structurally stable forms.Although their asymptotic behaviors to maintain memory are categorically distinct, their finite-time behaviors are similar.We build on the persistent manifold theory to explain the commonalities between bifurcations from and approximations of continuous attractors.Fast-slow decomposition analysis uncovers the existence of a persistent slow manifold that survives the seemingly destructive bifurcation, relating the flow within the manifold to the size of the perturbation. Moreover, this allows the bounding of the memory error of these approximations of continuous attractors.Finally, we train recurrent neural networks on analog memory tasks to support the appearance of these systems as solutions and their generalization capabilities.Therefore, we conclude that continuous attractors are functionally robust and remain useful as a universal analogy for understanding analog memory.

If there's one thing that turns people off from adopting new tech, it's being forced to look silly and feel uncomfortable for extended lengths of time. It was always the Achilles' heel for 3D in the past, and it remains the primary hurdle for VR headsets and goofy-looking smart glasses. Laptops, tablets, and even computer monitors have started embracing a new form of 3D technology that solves this problem entirely, without giving up just how compelling 3D can look. I've used the latest iteration of the technology and spoke with the creators--this might finally be the version of 3D that sticks. I was skeptical when I first saw this next generation of 3D technology. Interest in 3D comes in waves.

Pocketpair, the company behind last year's viral game Palworld, has a new venture: publishing indie games. Its first project, scheduled for release later this year, will be an as-yet-unnamed horror game from Surgent Studios, the developer behind 2024's Tales of Kenzera: Zau. Palworld, jokingly referred to as "Pokémon with guns," was a breakout success last year, drawing in more than 25 million players in its first few months. The company's step into publishing comes at a turbulent time for video games, especially smaller studios; last year, Among Us developer Innersloth announced its own move into publishing to help push projects forward. Pocketpair's Palworld success, it seems, is allowing it to do the same.

Consider a team of three guards (in green) trying to defend a fort from a team of three attackers (in red) in Figure 1. In this "Fort Attack" (FA) domain, each agent can move in one of four cardinal directions with a particular velocity, rotate clockwise or anticlockwise, shoot at an opponent within a given range, or do nothing. Each agent may have partial or full knowledge of the state of the world (e.g., location, status of each agent) at each step, but it has no prior experience of working with the other agents. Also, each agent may have limited (or no) ability to communicate with others. An episode ends when all members of a team are eliminated, an attacker reaches the fort, or the guards protect the fort for a sufficient time period.

In the last few months we have seen promising developments in establishing safeguards for AI. This includes a landmark EU regulation proposal on AI that prohibits unacceptable AI uses and imposes mandatory disclosures and evaluations for high-risk systems, an algorithmic transparency standard launched by the UK government, mandatory audits for AI hiring tech in New York City, and a draft AI Risk Assessment Framework developed by NIST at the request of US congress, to name a few. That being said, we are still in the early days of AI regulation. There is a long road ahead to minimize harms that algorithmic systems can cause. In this article series, I explore different topics related to the responsible use of AI and its societal implications.

The world of human artists will undoubtedly be shaken in the next months by this significant advancement for text-to-image AI models, for better or worse. And although I don't intend to bombard my email subscribers with AI image generator news, a lot of big updates are getting released every single day, I can't help but get amazed and share my excitement with you guys.

Embracer, the mega game publisher that's been snapping up new properties left and right, has made a deal to acquire the intellectual property catalogue and worldwide rights to various JRR Tolkien-related media and merch. To be precise, it will own the rights to "motion pictures, video games, board games, merchandising, theme parks and stage productions" based on the The Lord of the Rings trilogy and The Hobbit if the deal pushes through. It will also own the rights tied to any future literary work related to LOTR and The Hobbit that's authorized by the Tolkien Estate. This isn't the first Tolkien-related purchase Embracer has made: Back in 2021, it bought the board game publisher Asmodee, which has published over a dozen LOTR board games over the past 20 years. And if the acquisition goes through, Embracer will work with Amazon on The Lord of the Rings: Rings of Power series that will start streaming on September 2nd.