Random feature model with a nonlinear activation function has been shown to perform asymptotically equivalent to a Gaussian model in terms of training and generalization errors. Analysis of the equivalent model reveals an important yet not fully understood role played by the activation function. To address this issue, we study the "parameters" of the equivalent model to achieve improved generalization performance for a given supervised learning problem. We show that acquired parameters from the Gaussian model enable us to define a set of optimal nonlinearities. We provide two example classes from this set, e.g., second-order polynomial and piecewise linear functions. These functions are optimized to improve generalization performance regardless of the actual form. We experiment with regression and classification problems, including synthetic and real (e.g., CIFAR10) data. Our numerical results validate that the optimized nonlinearities achieve better generalization performance than widely-used nonlinear functions such as ReLU. Furthermore, we illustrate that the proposed nonlinearities also mitigate the so-called double descent phenomenon, which is known as the non-monotonic generalization performance regarding the sample size and the model size.

LEARNING a language used to mean sitting in a classroom and memorising hundreds of words. But thanks to apps like Duolingo, you can take a quick French lesson on your phone between meetings. Or you can use Google Translate to help you read French instead. As miraculous as these apps are, they don't work for everyone – especially if you speak an Indigenous language in regions like North America or New Zealand, where European settlers made …

Multimodal transfer learning aims to transform pretrained representations of diverse modalities into a common domain space for effective multimodal fusion. However, conventional systems are typically built on the assumption that all modalities exist, and the lack of modalities always leads to poor inference performance. Furthermore, extracting pretrained embeddings for all modalities is computationally inefficient for inference. In this work, to achieve high efficiency-performance multimodal transfer learning, we propose VideoAdviser, a video knowledge distillation method to transfer multimodal knowledge of video-enhanced prompts from a multimodal fundamental model (teacher) to a specific modal fundamental model (student). With an intuition that the best learning performance comes with professional advisers and smart students, we use a CLIP-based teacher model to provide expressive multimodal knowledge supervision signals to a RoBERTa-based student model via optimizing a step-distillation objective loss -- first step: the teacher distills multimodal knowledge of video-enhanced prompts from classification logits to a regression logit -- second step: the multimodal knowledge is distilled from the regression logit of the teacher to the student. We evaluate our method in two challenging multimodal tasks: video-level sentiment analysis (MOSI and MOSEI datasets) and audio-visual retrieval (VEGAS dataset). The student (requiring only the text modality as input) achieves an MAE score improvement of up to 12.3% for MOSI and MOSEI. Our method further enhances the state-of-the-art method by 3.4% mAP score for VEGAS without additional computations for inference. These results suggest the strengths of our method for achieving high efficiency-performance multimodal transfer learning.

Class-Incremental Learning (CIL) aims to build classification models from data streams. At each step of the CIL process, new classes must be integrated into the model. Due to catastrophic forgetting, CIL is particularly challenging when examples from past classes cannot be stored, the case on which we focus here. To date, most approaches are based exclusively on the target dataset of the CIL process. However, the use of models pre-trained in a self-supervised way on large amounts of data has recently gained momentum. The initial model of the CIL process may only use the first batch of the target dataset, or also use pre-trained weights obtained on an auxiliary dataset. The choice between these two initial learning strategies can significantly influence the performance of the incremental learning model, but has not yet been studied in depth. Performance is also influenced by the choice of the CIL algorithm, the neural architecture, the nature of the target task, the distribution of classes in the stream and the number of examples available for learning. We conduct a comprehensive experimental study to assess the roles of these factors. We present a statistical analysis framework that quantifies the relative contribution of each factor to incremental performance. Our main finding is that the initial training strategy is the dominant factor influencing the average incremental accuracy, but that the choice of CIL algorithm is more important in preventing forgetting. Based on this analysis, we propose practical recommendations for choosing the right initial training strategy for a given incremental learning use case. These recommendations are intended to facilitate the practical deployment of incremental learning.

When demonstrating a task, human tutors pedagogically modify their behavior by either "showing" the task rather than just "doing" it (exaggerating on relevant parts of the demonstration) or by giving demonstrations that best disambiguate the communicated goal. Analogously, human learners pragmatically infer the communicative intent of the tutor: they interpret what the tutor is trying to teach them and deduce relevant information for learning. Without such mechanisms, traditional Learning from Demonstration (LfD) algorithms will consider such demonstrations as sub-optimal. In this paper, we investigate the implementation of such mechanisms in a tutor-learner setup where both participants are artificial agents in an environment with multiple goals. Using pedagogy from the tutor and pragmatism from the learner, we show substantial improvements over standard learning from demonstrations.

We study unconstrained Online Linear Optimization with Lipschitz losses. The goal is to simultaneously achieve ($i$) second order gradient adaptivity; and ($ii$) comparator norm adaptivity also known as "parameter freeness" in the literature. Existing regret bounds (Cutkosky and Orabona, 2018; Mhammedi and Koolen, 2020; Jacobsen and Cutkosky, 2022) have the suboptimal $O(\sqrt{V_T\log V_T})$ dependence on the gradient variance $V_T$, while the present work improves it to the optimal rate $O(\sqrt{V_T})$ using a novel continuous-time-inspired algorithm, without any impractical doubling trick. This result can be extended to the setting with unknown Lipschitz constant, eliminating the range ratio problem from prior works (Mhammedi and Koolen, 2020). Concretely, we first show that the aimed simultaneous adaptivity can be achieved fairly easily in a continuous time analogue of the problem, where the environment is modeled by an arbitrary continuous semimartingale. Then, our key innovation is a new discretization argument that preserves such adaptivity in the discrete time adversarial setting. This refines a non-gradient-adaptive discretization argument from (Harvey et al., 2023), both algorithmically and analytically, which could be of independent interest.

This paper investigates fairness and bias in Canonical Correlation Analysis (CCA), a widely used statistical technique for examining the relationship between two sets of variables. We present a framework that alleviates unfairness by minimizing the correlation disparity error associated with protected attributes. Our approach enables CCA to learn global projection matrices from all data points while ensuring that these matrices yield comparable correlation levels to group-specific projection matrices. Experimental evaluation on both synthetic and real-world datasets demonstrates the efficacy of our method in reducing correlation disparity error without compromising CCA accuracy.

We present improved algorithms with worst-case regret guarantees for the stochastic linear bandit problem. The widely used "optimism in the face of uncertainty" principle reduces a stochastic bandit problem to the construction of a confidence sequence for the unknown reward function. The performance of the resulting bandit algorithm depends on the size of the confidence sequence, with smaller confidence sets yielding better empirical performance and stronger regret guarantees. In this work, we use a novel tail bound for adaptive martingale mixtures to construct confidence sequences which are suitable for stochastic bandits. These confidence sequences allow for efficient action selection via convex programming. We prove that a linear bandit algorithm based on our confidence sequences is guaranteed to achieve competitive worst-case regret. We show that our confidence sequences are tighter than competitors, both empirically and theoretically. Finally, we demonstrate that our tighter confidence sequences give improved performance in several hyperparameter tuning tasks.

The quickest way to second-guess a decision to major in English is this: have an extended family full of Salvadoran immigrants and pragmatic midwesterners. The ability to recite Chaucer in the original Middle English was unlikely to land me a job that would pay off my student loans and help me save for retirement, they suggested when I was a college freshman still figuring out my future. I stuck with English, but when my B.A. eventually spat me out into the thick of the Great Recession, I worried that they'd been right. After all, computer-science degrees, and certainly not English, have long been sold to college students as among the safest paths toward 21st-century job security. Coding jobs are plentiful across industries, and the pay is good--even after the tech layoffs of the past year.

Programmers often struggle to identify and fix bugs in their programs. In recent years, many language models (LMs) have been proposed to fix erroneous programs and support error recovery. However, the LMs tend to generate solutions that differ from the original input programs. This leads to potential comprehension difficulties for users. In this paper, we propose an approach to suggest a correct program with minimal repair edits using CodeT5. We fine-tune a pre-trained CodeT5 on code pairs of wrong and correct programs and evaluate its performance with several baseline models. The experimental results show that the fine-tuned CodeT5 achieves a pass@100 of 91.95% and an average edit distance of the most similar correct program of 6.84, which indicates that at least one correct program can be suggested by generating 100 candidate programs. We demonstrate the effectiveness of LMs in suggesting program repair with minimal edits for solving introductory programming problems.