Welcome to the Education Issue, our latest print magazine. It's becoming increasingly clear that we're in an entirely new place when it comes to the use of AI in education, and it is far from clear what that is going to mean. The world has changed, and there's no going back. Technologies like ChatGPT, OpenAI's massively mind-blowing generative AI software, will have all sorts of genuinely useful and transformative applications in the classroom. Yes, they will almost certainly also be used for cheating.

And then I did what an increasing number of us are doing: I turned to ChatGPT, OpenAI's massively mind-blowing generative AI software, to help me out. After training it on some of my previous work, I asked about the use of AI in education. AI is already doing big things in education. By crunching massive amounts of data on student performance, AI algorithms can tailor instruction to fit the needs of individual learners, which can mean big improvements in student outcomes. Chatbots and virtual assistants can provide students with on-the-spot assistance and feedback.

We seek to understand fundamental tradeoffs between the accuracy of prior information that a learner has on a given problem and its learning performance. We introduce the notion of prioritized risk, which differs from traditional notions of minimax and Bayes risk by allowing us to study such fundamental tradeoffs in settings where reality does not necessarily conform to the learner's prior. We present a general reduction-based approach for extending classical minimax lower-bound techniques in order to lower bound the prioritized risk for statistical estimation problems. We also introduce a novel generalization of Fano's inequality (which may be of independent interest) for lower bounding the prioritized risk in more general settings involving unbounded losses. We illustrate the ability of our framework to provide insights into tradeoffs between prior information and learning performance for problems in estimation, regression, and reinforcement learning.

Providing natural language instructions in prompts is a useful new paradigm for improving task performance of large language models in a zero-shot setting. Recent work has aimed to improve such prompts via manual rewriting or gradient-based tuning. However, manual rewriting is time-consuming and requires subjective interpretation, while gradient-based tuning can be extremely computationally demanding for large models and may not be feasible for API-based models. In this work, we introduce Gradient-free Instructional Prompt Search (GrIPS), a gradient-free, edit-based search approach for improving task instructions for large language models. GrIPS takes in instructions designed for humans and automatically returns an improved, edited prompt, while allowing for API-based tuning. With InstructGPT models, GrIPS improves the average task performance by up to 4.30 percentage points on eight classification tasks from the Natural Instructions dataset (with similar improvements for OPT, BLOOM, and FLAN-T5). We see improvements for both instruction-only prompts and instruction + k-shot examples prompts. Notably, GrIPS outperforms manual rewriting and purely example-based prompts while controlling for the available compute and data budget. Further, performance of GrIPS is comparable to select gradient-based tuning approaches. Qualitatively, we show our edits can simplify instructions and at times make them incoherent but nonetheless improve accuracy. Our code is available at: https://github.com/archiki/GrIPS

Deep reinforcement learning has achieved impressive results in recent years; yet, it is still severely troubled by environments showcasing sparse rewards. On top of that, not all sparse-reward environments are created equal; in other words, they can differ in the presence or absence of various features, with many of them having a great impact on learning. In light of this, the present work puts together a literature compilation of such environmental features, covering particularly those that have been taken advantage of and those that continue to pose a challenge. We expect this effort to provide guidance to researchers for assessing the generality of their new proposals and to call their attention to issues that remain unresolved when dealing with sparse rewards.

In this paper, we define, evaluate, and improve the "relay-generalization" performance of reinforcement learning (RL) agents on the out-of-distribution "controllable" states. Ideally, an RL agent that generally masters a task should reach its goal starting from any controllable state of the environment instead of memorizing a small set of trajectories. For example, a self-driving system should be able to take over the control from humans in the middle of driving and must continue to drive the car safely. To practically evaluate this type of generalization, we start the test agent from the middle of other independently well-trained stranger agents' trajectories. With extensive experimental evaluation, we show the prevalence of generalization failure on controllable states from stranger agents. For example, in the Humanoid environment, we observed that a well-trained Proximal Policy Optimization (PPO) agent, with only 3.9% failure rate during regular testing, failed on 81.6% of the states generated by well-trained stranger PPO agents. To improve "relay generalization," we propose a novel method called Self-Trajectory Augmentation (STA), which will reset the environment to the agent's old states according to the Q function during training. After applying STA to the Soft Actor Critic's (SAC) training procedure, we reduced the failure rate of SAC under relay-evaluation by more than three times in most settings without impacting agent performance and increasing the needed number of environment interactions. Our code is available at https://github.com/lan-lc/STA. Generalization is critical for deploying reinforcement learning (RL) agents into real-world applications. A well-trained RL agent that can achieve high rewards under restricted settings may not be able to handle the enormous state space and complex environment variations in the real world. There are many different aspects regarding the generalization of RL agents.

Recurrent neural networks trained with the backpropagation through time (BPTT) algorithm have led to astounding successes in various temporal tasks. However, BPTT introduces severe limitations, such as the requirement to propagate information backwards through time, the weight symmetry requirement, as well as update-locking in space and time. These problems become roadblocks for AI systems where online training capabilities are vital. Recently, researchers have developed biologically-inspired training algorithms, addressing a subset of those problems. In this work, we propose a novel learning algorithm called online spatio-temporal learning with target projection (OSTTP) that resolves all aforementioned issues of BPTT. In particular, OSTTP equips a network with the capability to simultaneously process and learn from new incoming data, alleviating the weight symmetry and update-locking problems. We evaluate OSTTP on two temporal tasks, showcasing competitive performance compared to BPTT. Moreover, we present a proof-of-concept implementation of OSTTP on a memristive neuromorphic hardware system, demonstrating its versatility and applicability to resource-constrained AI devices.

The rapid growth of social media has caused tremendous effects on information propagation, raising extreme challenges in detecting rumors. Existing rumor detection methods typically exploit the reposting propagation of a rumor candidate for detection by regarding all reposts to a rumor candidate as a temporal sequence and learning semantics representations of the repost sequence. However, extracting informative support from the topological structure of propagation and the influence of reposting authors for debunking rumors is crucial, which generally has not been well addressed by existing methods. In this paper, we organize a claim post in circulation as an adhoc event tree, extract event elements, and convert it to bipartite adhoc event trees in terms of both posts and authors, i.e., author tree and post tree. Accordingly, we propose a novel rumor detection model with hierarchical representation on the bipartite adhoc event trees called BAET. Specifically, we introduce word embedding and feature encoder for the author and post tree, respectively, and design a root-aware attention module to perform node representation. Then we adopt the tree-like RNN model to capture the structural correlations and propose a tree-aware attention module to learn tree representation for the author tree and post tree, respectively. Extensive experimental results on two public Twitter datasets demonstrate the effectiveness of BAET in exploring and exploiting the rumor propagation structure and the superior detection performance of BAET over state-of-the-art baseline methods.

Language models pre-trained on large self-supervised corpora, followed by task-specific fine-tuning has become the dominant paradigm in NLP. These pre-training datasets often have a one-to-many structure--e.g. in dialogue there are many valid responses for a given context. However, only some of these responses will be desirable in our downstream task. This raises the question of how we should train the model such that it can emulate the desirable behaviours, but not the undesirable ones. Current approaches train in a one-to-one setup--only a single target response is given for a single dialogue context--leading to models only learning to predict the average response, while ignoring the full range of possible responses. Using text-based games as a testbed, our approach, PASA, uses discrete latent variables to capture the range of different behaviours represented in our larger pre-training dataset. We then use knowledge distillation to distil the posterior probability distribution into a student model. This probability distribution is far richer than learning from only the hard targets of the dataset, and thus allows the student model to benefit from the richer range of actions the teacher model has learned. Results show up to 49% empirical improvement over the previous state-of-the-art model on the Jericho Walkthroughs dataset.

Pervasive computing allows the provision of services in many important areas, including the relevant and dynamic field of health and well-being. In this domain, Human Activity Recognition (HAR) has gained a lot of attention in recent years. Current solutions rely on Machine Learning (ML) models and achieve impressive results. However, the evolution of these models remains difficult, as long as a complete retraining is not performed. To overcome this problem, the concept of Continual Learning is very promising today and, more particularly, the techniques based on regularization. These techniques are particularly interesting for their simplicity and their low cost. Initial studies have been conducted and have shown promising outcomes. However, they remain very specific and difficult to compare. In this paper, we provide a comprehensive comparison of three regularization-based methods that we adapted to the HAR domain, highlighting their strengths and limitations. Our experiments were conducted on the UCI HAR dataset and the results showed that no single technique outperformed all others in all scenarios considered.