We consider a game-theoretical multi-agent learning problem where the feedback information can be lost during the learning process and rewards are given by a broad class of games known as variationally stable games. We propose a simple variant of the classical online gradient descent algorithm, called reweighted online gradient descent (ROGD) and show that in variationally stable games, if each agent adopts ROGD, then almost sure convergence to the set of Nash equilibria is guaranteed, even when the feedback loss is asynchronous and arbitrarily corrrelated among agents. We then extend the framework to deal with unknown feedback loss probabilities by using an estimator (constructed from past data) in its replacement. Finally, we further extend the framework to accomodate both asynchronous loss and stochastic rewards and establish that multi-agent ROGD learning still converges to the set of Nash equilibria in such settings. Together, these results contribute to the broad lanscape of multi-agent online learning by significantly relaxing the feedback information that is required to achieve desirable outcomes.

In the third season of "The Comeback," Kudrow has brought back her character Valerie Cherish, which had its roots at the Groundlings. A visitor to Stage 24 on the Warner Bros. lot, in Burbank, last November could be forgiven for thinking that the television show being filmed there was a sitcom called "How's That?!" The parking spaces outside were marked with "How's That?!" signs. Inside, director's chairs with the "How's That?!" logo were arranged around video monitors. The set--a New England bed-and-breakfast, with kitschy floral wallpaper--was surrounded by sitcom cameras and buzzing crew members wearing headsets. A studio audience filed into the bleachers, and a warmup comic urged them to "shake those funny bones." Then, with mounting gusto, he introduced the star of "How's That?!": "Here she is . . . the one and only . . . the living legend . . . She emerged to applause, in a potter's smock, wavy red hair under a bandanna, looking like a cross between Lucy Ricardo and Mrs. Garrett ...

Recently, learning discriminative features to improve the recognition performances gradually becomes the primary goal of deep learning, and numerous remarkable works have emerged. In this paper, we propose a novel yet extremely simple method Virtual Softmax to enhance the discriminative property of learned features by injecting a dynamic virtual negative class into the original softmax. Injecting virtual class aims to enlarge inter-class margin and compress intra-class distribution by strengthening the decision boundary constraint. Although it seems weird to optimize with this additional virtual class, we show that our method derives from an intuitive and clear motivation, and it indeed encourages the features to be more compact and separable. This paper empirically and experimentally demonstrates the superiority of Virtual Softmax, improving the performances on a variety of object classification and face verification tasks.

This month we tackle the topic of young people and what AI tools mean for their future. Joining the conversation this time are: Sanmay Das (Virginia Tech), Tom Dietterich (Oregon State University), Sabine Hauert (University of Bristol), Michael Littman (Brown University), and Ella Scallan (AIhub). As AI tools have become ubiquitous, we've seen growing concern and increasing coverage about how the use of such tools from a formative age might affect children. What do you think the impact will be and what skills might young people need to navigate this AI world? I met up with a bunch of high school friends when I was last in Switzerland and they were all wondering what their kids should study. They were wondering if they should do social science, seeing as AI tools have become adept at many tasks, such as coding, writing, art, etc. I think that we need social sciences, but that we also need people who know the technology and who can continue developing it. I say they should continue doing whatever they're interested in and those jobs will evolve and they'll look different, but there will still be a whole wealth of different types of jobs.

We study the following question: given an efficient approximation algorithm for an optimization problem, can we learn efficiently in the same setting? We give a formal affirmative answer to this question in the form of a reduction from online learning to offline approximate optimization using an efficient algorithm that guarantees near optimal regret. The algorithm is efficient in terms of the number of oracle calls to a given approximation oracle - it makes only logarithmically many such calls per iteration.

We design differentially private learning algorithms that are agnostic to the learning model assuming access to limited amount of unlabeled public data. First, we give a new differentially private algorithm for answering a sequence of $m$ online classification queries (given by a sequence of $m$ unlabeled public feature vectors) based on a private training set. Our private algorithm follows the paradigm of subsample-and-aggregate, in which any generic non-private learner is trained on disjoint subsets of the private training set, then for each classification query, the votes of the resulting classifiers ensemble are aggregated in a differentially private fashion.

Meta-learning for few-shot learning entails acquiring a prior over previous tasks and experiences, such that new tasks be learned from small amounts of data. However, a critical challenge in few-shot learning is task ambiguity: even when a powerful prior can be meta-learned from a large number of prior tasks, a small dataset for a new task can simply be too ambiguous to acquire a single model (e.g., a classifier) for that task that is accurate. In this paper, we propose a probabilistic meta-learning algorithm that can sample models for a new task from a model distribution. Our approach extends model-agnostic meta-learning, which adapts to new tasks via gradient descent, to incorporate a parameter distribution that is trained via a variational lower bound. At meta-test time, our algorithm adapts via a simple procedure that injects noise into gradient descent, and at meta-training time, the model is trained such that this stochastic adaptation procedure produces samples from the approximate model posterior. Our experimental results show that our method can sample plausible classifiers and regressors in ambiguous few-shot learning problems. We also show how reasoning about ambiguity can also be used for downstream active learning problems.

We don't know if AI-powered toys are safe, but they're here anyway Toys powered by AI show a worrying lack of emotional understanding. Mya, aged 3, and her mother Vicky playing with an AI toy called Gabbo during an observation at the University of Cambridge's Faculty of Education Even the most cutting-edge AI models are prone to presenting fabrication as fact, dispensing dangerous information and failing to grasp social cues. Despite this, toys equipped with AI that can chat with children are a burgeoning industry. Some scientists are warning that the devices could be risky and require strict regulation. In the latest study, researchers even observed a 5-year-old telling such a toy "I love you", to which it replied: "As a friendly reminder, please ensure interactions adhere to the guidelines provided. Let me know how you would like to proceed."

Claire chatted to Jamie Palmer from Icarus Robotics about building a robotic labour force to perform routine and risky tasks in orbit. Jamie Palmer is co-founder and CTO of Icarus Robotics . He earned a Master's in Robotics from Columbia University on a full scholarship, researching intelligent, dexterous manipulation in the ROAM lab. Jamie developed and deployed autonomous hospital robots during the pandemic and worked as a race-winning engineer for the Mercedes-AMG Petronas Formula One team. Robot Talk is a weekly podcast that explores the exciting world of robotics, artificial intelligence and autonomous machines.

In this interview series, we're meeting some of the AAAI/SIGAI Doctoral Consortium participants to find out more about their research. We caught up with Oliver Chang whose research interests span deep reinforcement learning, autonomous vehicles, and explainable AI. We found out more about some of the projects he's worked on so far, what drew him to the field, and what future AI directions he's excited about. Could you give us a quick introduction to who you are, where you're studying, and the topic of your research? I'm specializing in reinforcement learning applied to autonomous vehicles and UAVs.