Quantum computing and machine learning are two of the most exciting technologies that can transform businesses. We can only imagine how powerful it can be if we can combine the power of both of these technologies. When we can integrate quantum algorithms in programs based on machine learning, that is called quantum machine learning. This fascinating area has been a major area of tech firms, and they have brought out tools and platforms to deploy such algorithms effectively. Some of these include TensorFlow Quantum from Google, Quantum Machine Learning (QML) library from Microsoft, QC Ware Forge built on Amazon Braket, etc. Students skilled in working with quantum machine learning algorithms can be in great demand due to the opportunities the field holds.

This March, IES Director Mark Schneider released a blog in which he discussed exploring a partnership with the National Science Foundation (NSF) to encourage scientists with expertise in AI and related fields to address the important post-pandemic need for accelerating learning. IES is now excited to announce our resulting participation in NSF's National Artificial Intelligence (AI) Research Institutes--Accelerating Research, Transforming Society, and Growing the American Workforce solicitation. In this blog, we describe this new funding opportunity, provide examples of existing NCSER-funded research in this area, and highlight the potential for such research to further improve outcomes for learners with disabilities. With funding from the American Rescue Plan, NCSER plans to support research under Theme 6, Track B: AI-Augmented Learning for Individuals with Disabilities. Proposals must discuss how the work will respond to the needs of learners with or at risk for a disability in an area where the COVID-19 pandemic has further widened existing gaps and/or resulted in decreased access and opportunities for students with disabilities to learn and receive support services.

State-of-the-art sign language generation frameworks lack expressivity and naturalness which is the result of only focusing manual signs, neglecting the affective, grammatical and semantic functions of facial expressions. The purpose of this work is to augment semantic representation of sign language through grounding facial expressions. We study the effect of modeling the relationship between text, gloss, and facial expressions on the performance of the sign generation systems. In particular, we propose a Dual Encoder Transformer able to generate manual signs as well as facial expressions by capturing the similarities and differences found in text and sign gloss annotation. We take into consideration the role of facial muscle activity to express intensities of manual signs by being the first to employ facial action units in sign language generation. We perform a series of experiments showing that our proposed model improves the quality of automatically generated sign language.

When people receive advice while making difficult decisions, they often make better decisions in the moment and also increase their knowledge in the process. However, such incidental learning can only occur when people cognitively engage with the information they receive and process this information thoughtfully. How do people process the information and advice they receive from AI, and do they engage with it deeply enough to enable learning? To answer these questions, we conducted three experiments in which individuals were asked to make nutritional decisions and received simulated AI recommendations and explanations. In the first experiment, we found that when people were presented with both a recommendation and an explanation before making their choice, they made better decisions than they did when they received no such help, but they did not learn. In the second experiment, participants first made their own choice, and only then saw a recommendation and an explanation from AI; this condition also resulted in improved decisions, but no learning. However, in our third experiment, participants were presented with just an AI explanation but no recommendation and had to arrive at their own decision. This condition led to both more accurate decisions and learning gains. We hypothesize that learning gains in this condition were due to deeper engagement with explanations needed to arrive at the decisions. This work provides some of the most direct evidence to date that it may not be sufficient to include explanations together with AI-generated recommendation to ensure that people engage carefully with the AI-provided information. This work also presents one technique that enables incidental learning and, by implication, can help people process AI recommendations and explanations more carefully.

In the quest for autonomous agents learning open-ended repertoires of skills, most works take a Piagetian perspective: learning trajectories are the results of interactions between developmental agents and their physical environment. The Vygotskian perspective, on the other hand, emphasizes the centrality of the socio-cultural environment: higher cognitive functions emerge from transmissions of socio-cultural processes internalized by the agent. This paper argues that both perspectives could be coupled within the learning of autotelic agents to foster their skill acquisition. To this end, we make two contributions: 1) a novel social interaction protocol called Help Me Explore (HME), where autotelic agents can benefit from both individual and socially guided exploration. In social episodes, a social partner suggests goals at the frontier of the learning agent knowledge. In autotelic episodes, agents can either learn to master their own discovered goals or autonomously rehearse failed social goals; 2) GANGSTR, a graph-based autotelic agent for manipulation domains capable of decomposing goals into sequences of intermediate sub-goals. We show that when learning within HME, GANGSTR overcomes its individual learning limits by mastering the most complex configurations (e.g. stacks of 5 blocks) with only few social interventions.

We present LaMDA: Language Models for Dialog Applications. LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters and are pre-trained on 1.56T words of public dialog data and web text. While model scaling alone can improve quality, it shows less improvements on safety and factual grounding. We demonstrate that fine-tuning with annotated data and enabling the model to consult external knowledge sources can lead to significant improvements towards the two key challenges of safety and factual grounding. The first challenge, safety, involves ensuring that the model's responses are consistent with a set of human values, such as preventing harmful suggestions and unfair bias. We quantify safety using a metric based on an illustrative set of human values, and we find that filtering candidate responses using a LaMDA classifier fine-tuned with a small amount of crowdworker-annotated data offers a promising approach to improving model safety. The second challenge, factual grounding, involves enabling the model to consult external knowledge sources, such as an information retrieval system, a language translator, and a calculator. We quantify factuality using a groundedness metric, and we find that our approach enables the model to generate responses grounded in known sources, rather than responses that merely sound plausible. Finally, we explore the use of LaMDA in the domains of education and content recommendations, and analyze their helpfulness and role consistency.

In this paper, we consider learning scenarios where the learned model is evaluated under an unknown test distribution which potentially differs from the training distribution (i.e. distribution shift). The learner has access to a family of weight functions such that the test distribution is a reweighting of the training distribution under one of these functions, a setting typically studied under the name of Distributionally Robust Optimization (DRO). We consider the problem of deriving regret bounds in the classical learning theory setting, and require that the resulting regret bounds hold uniformly for all potential test distributions. We show that the DRO formulation does not guarantee uniformly small regret under distribution shift. We instead propose an alternative method called Minimax Regret Optimization (MRO), and show that under suitable conditions this method achieves uniformly low regret across all test distributions. We also adapt our technique to have stronger guarantees when the test distributions are heterogeneous in their similarity to the training data. Given the widespead optimization of worst case risks in current approaches to robust machine learning, we believe that MRO can be a strong alternative to address distribution shift scenarios.

Large-scale machine learning systems often involve data distributed across a collection of users. Federated optimization algorithms leverage this structure by communicating model updates to a central server, rather than entire datasets. In this paper, we study stochastic optimization algorithms for a personalized federated learning setting involving local and global models subject to user-level (joint) differential privacy. While learning a private global model induces a cost of privacy, local learning is perfectly private. We show that coordinating local learning with private centralized learning yields a generically useful and improved tradeoff between accuracy and privacy. We illustrate our theoretical results with experiments on synthetic and real-world datasets.

Online selection of optimal waveforms for target tracking with active sensors has long been a problem of interest. Many conventional solutions utilize an estimation-theoretic interpretation, in which a waveform-specific Cram\'{e}r-Rao lower bound on measurement error is used to select the optimal waveform for each tracking step. However, this approach is only valid in the high SNR regime, and requires a rather restrictive set of assumptions regarding the target motion and measurement models. Further, due to computational concerns, many traditional approaches are limited to near-term, or myopic, optimization, even though radar scenes exhibit strong temporal correlation. More recently, reinforcement learning has been proposed for waveform selection, in which the problem is framed as a Markov decision process (MDP), allowing for long-term planning. However, a major limitation of reinforcement learning is that the memory length of the underlying Markov process is often unknown for realistic target and channel dynamics, and a more general framework is desirable. This work develops a universal sequential waveform selection scheme which asymptotically achieves Bellman optimality in any radar scene which can be modeled as a $U^{\text{th}}$ order Markov process for a finite, but unknown, integer $U$. Our approach is based on well-established tools from the field of universal source coding, where a stationary source is parsed into variable length phrases in order to build a context-tree, which is used as a probabalistic model for the scene's behavior. We show that an algorithm based on a multi-alphabet version of the Context-Tree Weighting (CTW) method can be used to optimally solve a broad class of waveform-agile tracking problems while making minimal assumptions about the environment's behavior.

We give exponential statistical query (SQ) lower bounds for learning two-hidden-layer ReLU networks with respect to Gaussian inputs in the standard (noise-free) model. No general SQ lower bounds were known for learning ReLU networks of any depth in this setting: previous SQ lower bounds held only for adversarial noise models (agnostic learning) or restricted models such as correlational SQ. Prior work hinted at the impossibility of our result: Vempala and Wilmes showed that general SQ lower bounds cannot apply to any real-valued family of functions that satisfies a simple non-degeneracy condition. To circumvent their result, we refine a lifting procedure due to Daniely and Vardi that reduces Boolean PAC learning problems to Gaussian ones. We show how to extend their technique to other learning models and, in many well-studied cases, obtain a more efficient reduction. As such, we also prove new cryptographic hardness results for PAC learning two-hidden-layer ReLU networks, as well as new lower bounds for learning constant-depth ReLU networks from membership queries.