Generative artificial intelligence (GenAI) is increasingly becoming a part of work practices across the technology industry and being used across a range of industries. This has necessitated the need to better understand how GenAI is being used by professionals in the field so that we can better prepare students for the workforce. An improved understanding of the use of GenAI in practice can help provide guidance on the design of GenAI literacy efforts including how to integrate it within courses and curriculum, what aspects of GenAI to teach, and even how to teach it. This paper presents a field study that compares the use of GenAI across three different functions - product development, software engineering, and digital content creation - to identify how GenAI is currently being used in the industry. This study takes a human augmentation approach with a focus on human cognition and addresses three research questions: how is GenAI augmenting work practices; what knowledge is important and how are workers learning; and what are the implications for training the future workforce. Findings show a wide variance in the use of GenAI and in the level of computing knowledge of users. In some industries GenAI is being used in a highly technical manner with deployment of fine-tuned models across domains. Whereas in others, only off-the-shelf applications are being used for generating content. This means that the need for what to know about GenAI varies, and so does the background knowledge needed to utilize it. For the purposes of teaching and learning, our findings indicated that different levels of GenAI understanding needs to be integrated into courses. From a faculty perspective, the work has implications for training faculty so that they are aware of the advances and how students are possibly, as early adopters, already using GenAI to augment their learning practices.

In imitation learning for robotic manipulation, decomposing object manipulation tasks into multiple sub-tasks is essential. This decomposition enables the reuse of learned skills in varying contexts and the combination of acquired skills to perform novel tasks, rather than merely replicating demonstrated motions. Gaze plays a critical role in human object manipulation, where it is strongly correlated with hand movements. We hypothesize that an imitating agent's gaze control, fixating on specific landmarks and transitioning between them, simultaneously segments demonstrated manipulations into sub-tasks. In this study, we propose a simple yet robust task decomposition method based on gaze transitions. The method leverages teleoperation, a common modality in robotic manipulation for collecting demonstrations, in which a human operator's gaze is measured and used for task decomposition as a substitute for an imitating agent's gaze. Notably, our method achieves consistent task decomposition across all demonstrations for each task, which is desirable in contexts such as machine learning. We applied this method to demonstrations of various tasks and evaluated the characteristics and consistency of the resulting sub-tasks. Furthermore, through extensive testing across a wide range of hyperparameter variations, we demonstrated that the proposed method possesses the robustness necessary for application to different robotic systems.

The success of deep reinforcement learning (DRL) relies on the availability and quality of training data, often requiring extensive interactions with specific environments. In many real-world scenarios, where data collection is costly and risky, offline reinforcement learning (RL) offers a solution by utilizing data collected by domain experts and searching for a batch-constrained optimal policy. This approach is further augmented by incorporating external data sources, expanding the range and diversity of data collection possibilities. However, existing offline RL methods often struggle with challenges posed by non-matching data from these external sources. In this work, we specifically address the problem of source-target domain mismatch in scenarios involving mixed datasets, characterized by a predominance of source data generated from random or suboptimal policies and a limited amount of target data generated from higher-quality policies. To tackle this problem, we introduce Transition Scoring (TS), a novel method that assigns scores to transitions based on their similarity to the target domain, and propose Curriculum Learning-Based Trajectory Valuation (CLTV), which effectively leverages these transition scores to identify and prioritize high-quality trajectories through a curriculum learning approach. Our extensive experiments across various offline RL methods and MuJoCo environments, complemented by rigorous theoretical analysis, demonstrate that CLTV enhances the overall performance and transferability of policies learned by offline RL algorithms.

The main goal of this article is to convince you, the reader, that supervised learning in the Probably Approximately Correct (PAC) model is closely related to -- of all things -- bipartite matching! En-route from PAC learning to bipartite matching, I will overview a particular transductive model of learning, and associated one-inclusion graphs, which can be viewed as a generalization of some of the hat puzzles that are popular in recreational mathematics. Whereas this transductive model is far from new, it has recently seen a resurgence of interest as a tool for tackling deep questions in learning theory. A secondary purpose of this article could be as a (biased) tutorial on the connections between the PAC and transductive models of learning.

Self-distillation (SD), a technique where a model refines itself from its own predictions, has garnered attention as a simple yet powerful approach in machine learning. Despite its widespread use, the mechanisms underlying its effectiveness remain unclear. In this study, we investigate the efficacy of hyperparameter-tuned multi-stage SD in binary classification tasks with noisy labeled Gaussian mixture data, utilizing a replica theory. Our findings reveals that the primary driver of SD's performance improvement is denoising through hard pseudo-labels, with the most notable gains observed in moderately sized datasets. We also demonstrate the efficacy of practical heuristics, such as early stopping for extracting meaningful signal and bias fixation for imbalanced data. These results provide both theoretical guarantees and practical insights, advancing our understanding and application of SD in noisy settings.

So you want to learn German. Are you going to watch a bunch of YouTube videos and try to brute force your way to fluency? Learning a programming language like Python is a little different from trying to speak a new language, but learning it on your own is still a major challenge. That's why it's useful to check out resources like the Complete Python Certification Boot Camp. This 12-course meal of knowledge is a beginner-friendly way to learn Python, and it's on sale for 19.99.

Deep Neural Networks (DNNs) excel at many tasks, often rivaling or surpassing human performance. Yet their internal processes remain elusive, frequently described as "black boxes." While performance can be refined experimentally, achieving a fundamental grasp of their inner workings is still a challenge. Statistical Mechanics has long tackled computational problems, and this thesis applies physics-based insights to understand DNNs via three complementary approaches. First, by averaging over data, we derive an asymptotic bound on generalization that depends solely on the size of the last layer, rather than on the total number of parameters -- revealing how deep architectures process information differently across layers. Second, adopting a data-dependent viewpoint, we explore a finite-width thermodynamic limit beyond the infinite-width regime. This leads to: (i) a closed-form expression for the generalization error in a finite-width one-hidden-layer network (regression task); (ii) an approximate partition function for deeper architectures; and (iii) a link between deep networks in this thermodynamic limit and Student's t-processes. Finally, from a task-explicit perspective, we present a preliminary analysis of how DNNs interact with a controlled dataset, investigating whether they truly internalize its structure -- collapsing to the teacher -- or merely memorize it. By understanding when a network must learn data structure rather than just memorize, it sheds light on fostering meaningful internal representations. In essence, this thesis leverages the synergy between Statistical Physics and Machine Learning to illuminate the inner behavior of DNNs.

We show that learning-rate schedules for large model training behave surprisingly similar to a performance bound from non-smooth convex optimization theory. We provide a bound for the constant schedule with linear cooldown; in particular, the practical benefit of cooldown is reflected in the bound due to the absence of logarithmic terms. Further, we show that this surprisingly close match between optimization theory and practice can be exploited for learning-rate tuning: we achieve noticeable improvements for training 124M and 210M Llama-type models by (i) extending the schedule for continued training with optimal learning-rate, and (ii) transferring the optimal learning-rate across schedules.

Cold-start active learning (CSAL) selects valuable instances from an unlabeled dataset for manual annotation. It provides high-quality data at a low annotation cost for label-scarce text classification. However, existing CSAL methods overlook weak classes and hard representative examples, resulting in biased learning. To address these issues, this paper proposes a novel dual-diversity enhancing and uncertainty-aware (DEUCE) framework for CSAL. Specifically, DEUCE leverages a pretrained language model (PLM) to efficiently extract textual representations, class predictions, and predictive uncertainty. Then, it constructs a Dual-Neighbor Graph (DNG) to combine information on both textual diversity and class diversity, ensuring a balanced data distribution. It further propagates uncertainty information via density-based clustering to select hard representative instances. DEUCE performs well in selecting class-balanced and hard representative data by dual-diversity and informativeness. Experiments on six NLP datasets demonstrate the superiority and efficiency of DEUCE.

The notion of generalization has moved away from the classical one defined in statistical learning theory towards an emphasis on out-of-domain generalization (OODG). Recently, there is a growing focus on inductive generalization, where a progression of difficulty implicitly governs the direction of domain shifts. In inductive generalization, it is often assumed that the training data lie in the easier side, while the testing data lie in the harder side. The challenge is that training data are always finite, but a learner is expected to infer an inductive principle that could be applied in an unbounded manner. This emerging regime has appeared in the literature under different names, such as length/logical/algorithmic extrapolation, but a formal definition is lacking. This work provides such a formalization that centers on the concept of model successors. Then we outline directions to adapt well-established techniques towards the learning of model successors. This work calls for restructuring of the research discussion around inductive generalization from fragmented task-centric communities to a more unified effort, focused on universal properties of learning and computation.