Since the representative capacity of graph-based clustering methods is usually limited by the graph constructed on the original features, it is attractive to find whether graph neural networks (GNNs) can be applied to augment the capacity. The core problems mainly come from two aspects: (1) the graph is unavailable in the most clustering scenes so that how to construct high-quality graphs on the non-graph data is usually the most important part; (2) given n samples, the graph-based clustering methods usually consume at least $\mathcal O(n^2)$ time to build graphs and the graph convolution requires nearly $\mathcal O(n^2)$ for a dense graph and $\mathcal O(|\mathcal{E}|)$ for a sparse one with $|\mathcal{E}|$ edges. Accordingly, both graph-based clustering and GNNs suffer from the severe inefficiency problem. To tackle these problems, we propose a novel clustering method, AnchorGAE, with the self-supervised estimation of graph and efficient graph convolution. We first show how to convert a non-graph dataset into a graph dataset, by introducing the generative graph model and anchors. We then show that the constructed bipartite graph can reduce the computational complexity of graph convolution from $\mathcal O(n^2)$ and $\mathcal O(|\mathcal{E}|)$ to $\mathcal O(n)$. The succeeding steps for clustering can be easily designed as $\mathcal O(n)$ operations. Interestingly, the anchors naturally lead to siamese architecture with the help of the Markov process. Furthermore, the estimated bipartite graph is updated dynamically according to the features extracted by GNN, to promote the quality of the graph. However, we theoretically prove that the self-supervised paradigm frequently results in a collapse that often occurs after 2-3 update iterations in experiments, especially when the model is well-trained. A specific strategy is accordingly designed to prevent the collapse.

Language models have recently achieved strong performance across a wide range of NLP benchmarks. However, unlike benchmarks, real world tasks are often poorly specified, and agents must deduce the user's intended behavior from a combination of context, instructions, and examples. We investigate how both humans and models behave in the face of such task ambiguity by proposing AmbiBench, a new benchmark of six ambiguously-specified classification tasks. We evaluate humans and models on AmbiBench by seeing how well they identify the intended task using 1) instructions with varying degrees of ambiguity, and 2) different numbers of labeled examples. We find that the combination of model scaling (to 175B parameters) and training with human feedback data enables models to approach or exceed the accuracy of human participants across tasks, but that either one alone is not sufficient. In addition, we show how to dramatically improve the accuracy of language models trained without large-scale human feedback training by finetuning on a small number of ambiguous in-context examples, providing a promising direction for teaching models to generalize well in the face of ambiguity. Language models have recently been applied to a wide range of NLP benchmarks, ranging from question answering, summarization, and logical reasoning, to solving riddles, dark humor detection, and ASCII word recognition (Brown et al., 2020; Srivastava et al., 2022). Performance across tasks has improved as models and datasets have grown in size, raising the prospect of a route towards generalist NLP models with broad utility. However, one feature many of these benchmarks share is that they are carefully designed to make the desired task very clear to the language model, since this is a prerequisite for establishing performance on that task. Figure 1: Complex tasks are often hard to specify precisely, leaving important pieces of information missing. Agents should be able to fill in the blanks by combining information from instructions and examples in order to identify the intended behavior. Rather than iterating over and perfecting a specification for their tasks, everyday users of language models may wish to define tasks on an as-needed basis, without worrying that they will be misunderstood. This is especially important for safe and robust deployment of language models, as such undesirable dependencies can be hidden hazards that are only revealed when a model fails catastrophically in a new setting (Geirhos et al., 2020). To operationalize this problem, we introduce AmbiBench, a new benchmark of six ambiguouslyspecified tasks.

We explore the use of knowledge distillation (KD) for learning compact and accurate models that enable classification of animal behavior from accelerometry data on wearable devices. To this end, we take a deep and complex convolutional neural network, known as residual neural network (ResNet), as the teacher model. ResNet is specifically designed for multivariate time-series classification. We use ResNet to distill the knowledge of animal behavior classification datasets into soft labels, which consist of the predicted pseudo-probabilities of every class for each datapoint. We then use the soft labels to train our significantly less complex student models, which are based on the gated recurrent unit (GRU) and multilayer perceptron (MLP). The evaluation results using two real-world animal behavior classification datasets show that the classification accuracy of the student GRU-MLP models improves appreciably through KD, approaching that of the teacher ResNet model. To further reduce the computational and memory requirements of performing inference using the student models trained via KD, we utilize dynamic fixed-point quantization (DQ) through an appropriate modification of the computational graph of the considered models. We implement both unquantized and quantized versions of the developed KD-based models on the embedded systems of our purpose-built collar and ear tag devices to classify animal behavior in situ and in real time. Our evaluations corroborate the effectiveness of KD and DQ in improving the accuracy and efficiency of in-situ animal behavior classification.

Humans can leverage physical interaction to teach robot arms. As the human kinesthetically guides the robot through demonstrations, the robot learns the desired task. While prior works focus on how the robot learns, it is equally important for the human teacher to understand what their robot is learning. Visual displays can communicate this information; however, we hypothesize that visual feedback alone misses out on the physical connection between the human and robot. In this paper we introduce a novel class of soft haptic displays that wrap around the robot arm, adding signals without affecting that interaction. We first design a pneumatic actuation array that remains flexible in mounting. We then develop single and multi-dimensional versions of this wrapped haptic display, and explore human perception of the rendered signals during psychophysic tests and robot learning. We ultimately find that people accurately distinguish single-dimensional feedback with a Weber fraction of 11.4%, and identify multi-dimensional feedback with 94.5% accuracy. When physically teaching robot arms, humans leverage the single- and multi-dimensional feedback to provide better demonstrations than with visual feedback: our wrapped haptic display decreases teaching time while increasing demonstration quality. This improvement depends on the location and distribution of the wrapped haptic display. You can see videos of our device and experiments here: https://youtu.be/yPcMGeqsjdM

Taking advice from others requires confidence in their competence. This is important for interaction with peers, but also for collaboration with social robots and artificial agents. Nonetheless, we do not always have access to information about others' competence or performance. In these uncertain environments, do our prior beliefs about the nature and the competence of our interacting partners modulate our willingness to rely on their judgments? In a joint perceptual decision making task, participants made perceptual judgments and observed the simulated estimates of either a human participant, a social humanoid robot or a computer. Then they could modify their estimates based on this feedback. Results show participants' belief about the nature of their partner biased their compliance with its judgments: participants were more influenced by the social robot than human and computer partners. This difference emerged strongly at the very beginning of the task and decreased with repeated exposure to empirical feedback on the partner's responses, disclosing the role of prior beliefs in social influence under uncertainty. Furthermore, the results of our functional task suggest an important difference between human-human and human-robot interaction in the absence of overt socially relevant signal from the partner: the former is modulated by social normative mechanisms, whereas the latter is guided by purely informational mechanisms linked to the perceived competence of the partner.

This AI technology specializes in dialogue and generates original responses to questions in moments. The answers are sophisticated, structured, and in the words of one high school teacher, are "better than the large majority of writing seen by your average teacher or professor." The implications of AI technology for education have many educators concerned that this signals the "end of English" and the end of writing as an authentic assessment. I understand why a technology like ChatGPT elicits such a powerful response from educators. It threatens to disrupt the approach many schools and colleges use to teach students.

We need career planners, and not just people to get admission! Thanks to the pandemic, which pushed the limits of online aggregators and EdTech companies in international student recruitment. We are witnessing technology slowly making a powerful impact on student recruitment; though the industry is yet to witness the full power of Artificial Intelligence and Automation. Building AI platforms are going to be cheaper and replicating a technology model doesn't require a big innovation. The industry is going to be dumped with too much data for recruiters and students.

Raphael Zufferey, a postdoctoral fellow in the Laboratory of Intelligent Systems (LIS) and Biorobotics ab (BioRob) in the School of Engineering, is the first author on a recent Nature Communications paper describing the unique landing gear that makes such perching possible. He built and tested it in collaboration with colleagues at the University of Seville, Spain, where the 700-gram ornithopter itself was developed as part of the European project GRIFFIN. "This is the first phase of a larger project. Once an ornithopter can master landing autonomously on a tree branch, then it has the potential to carry out specific tasks, such as unobtrusively collecting biological samples or measurements from a tree. Eventually, it could even land on artificial structures, which could open up further areas of application," Zufferey says.

Bringing a machine learning model into the real world involves a lot more than just modeling. This Specialization will teach you how to navigate various deployment scenarios and use data more effectively to train your model. In this third course, you will: - Perform streamlined ETL tasks using TensorFlow Data Services - Load different datasets and custom feature vectors using TensorFlow Hub and TensorFlow Data Services APIs - Create and use pre-built pipelines for generating highly reproducible I/O pipelines for any dataset - Optimize data pipelines that become a bottleneck in the training process - Publish your own datasets to the TensorFlow Hub library and share standardized data with researchers and developers around the world This Specialization builds upon our TensorFlow in Practice Specialization. If you are new to TensorFlow, we recommend that you take the TensorFlow in Practice Specialization first. To develop a deeper, foundational understanding of how neural networks work, we recommend that you take the Deep Learning Specialization.

When I tell people that I work on getting robots to cook and do household chores, they often ask me why this is so difficult. "A child can learn to make an omelet," they say. "Why is it so hard for a robot?" I usually tell them that they think it's so easy because they're thinking like a human. To a machine, making an omelet is a lofty, long-horizon goal.