亚洲男人的天堂2018av,欧美草比,久久久久久免费视频精选,国色天香在线看免费,久久久久亚洲av成人片仓井空

Gaussianization is a simple generative model that can be trained without backpropagation. It has shown compelling performance on low dimensional data. As the dimension increases, however, it has been observed that the convergence speed slows down. We show analytically that the number of required layers scales linearly with the dimension for Gaussian input. We argue that this is because the model is unable to capture dependencies between dimensions. Empirically, we find the same linear increase in cost for arbitrary input $p(x)$, but observe favorable scaling for some distributions. We explore potential speed-ups and formulate challenges for further research.

In this study, we focus on learning Hamiltonian systems, which involves predicting the coordinate (q) and momentum (p) variables generated by a symplectic mapping. Based on Chen & Tao (2021), the symplectic mapping is represented by a generating function. To extend the prediction time period, we develop a new learning scheme by splitting the time series (q_i, p_i) into several partitions. We then train a large-step neural network (LSNN) to approximate the generating function between the first partition (i.e. the initial condition) and each one of the remaining partitions. This partition approach makes our LSNN effectively suppress the accumulative error when predicting the system evolution. Then we train the LSNN to learn the motions of the 2:3 resonant Kuiper belt objects for a long time period of 25000 yr. The results show that there are two significant improvements over the neural network constructed in our previous work (Li et al. 2022): (1) the conservation of the Jacobi integral, and (2) the highly accurate predictions of the orbital evolution. Overall, we propose that the designed LSNN has the potential to considerably improve predictions of the long-term evolution of more general Hamiltonian systems.

Automatic code optimization is a complex process that typically involves the application of multiple discrete algorithms that modify the program structure irreversibly. However, the design of these algorithms is often monolithic, and they require repetitive implementation to perform similar analyses due to the lack of cooperation. To address this issue, modern optimization techniques, such as equality saturation, allow for exhaustive term rewriting at various levels of inputs, thereby simplifying compiler design. In this paper, we propose equality saturation to optimize sequential codes utilized in directive-based programming for GPUs. Our approach simultaneously realizes less computation, less memory access, and high memory throughput. Our fully-automated framework constructs single-assignment forms from inputs to be entirely rewritten while keeping dependencies and extracts optimal cases. Through practical benchmarks, we demonstrate a significant performance improvement on several compilers. Furthermore, we highlight the advantages of computational reordering and emphasize the significance of memory-access order for modern GPUs.

Spiking neural networks (SNN) are able to learn spatiotemporal features while using less energy, especially on neuromorphic hardware. The most widely used spiking neuron in deep learning is the Leaky Integrate and Fire (LIF) neuron. LIF neurons operate sequentially, however, since the computation of state at time t relies on the state at time t-1 being computed. This limitation is shared with Recurrent Neural Networks (RNN) and results in slow training on Graphics Processing Units (GPU). In this paper, we propose the Stochastic Parallelizable Spiking Neuron (SPSN) to overcome the sequential training limitation of LIF neurons. By separating the linear integration component from the non-linear spiking function, SPSN can be run in parallel over time. The proposed approach results in performance comparable with the state-of-the-art for feedforward neural networks on the Spiking Heidelberg Digits (SHD) dataset, outperforming LIF networks while training 10 times faster and outperforming non-spiking networks with the same network architecture. For longer input sequences of 10000 time-steps, we show that the proposed approach results in 4000 times faster training, thus demonstrating the potential of the proposed approach to accelerate SNN training for very large datasets.

Following the traditional paradigm of convolutional neural networks (CNNs), modern CNNs manage to keep pace with more recent, for example transformer-based, models by not only increasing model depth and width but also the kernel size. This results in large amounts of learnable model parameters that need to be handled during training. While following the convolutional paradigm with the according spatial inductive bias, we question the significance of \emph{learned} convolution filters. In fact, our findings demonstrate that many contemporary CNN architectures can achieve high test accuracies without ever updating randomly initialized (spatial) convolution filters. Instead, simple linear combinations (implemented through efficient $1\times 1$ convolutions) suffice to effectively recombine even random filters into expressive network operators. Furthermore, these combinations of random filters can implicitly regularize the resulting operations, mitigating overfitting and enhancing overall performance and robustness. Conversely, retaining the ability to learn filter updates can impair network performance. Lastly, although we only observe relatively small gains from learning $3\times 3$ convolutions, the learning gains increase proportionally with kernel size, owing to the non-idealities of the independent and identically distributed (\textit{i.i.d.}) nature of default initialization techniques.

Learning-based solutions for vision tasks require a large amount of labeled training data to ensure their performance and reliability. In single-task vision-based settings, inconsistency-based active learning has proven to be effective in selecting informative samples for annotation. However, there is a lack of research exploiting the inconsistency between multiple tasks in multi-task networks. To address this gap, we propose a novel multi-task active learning strategy for two coupled vision tasks: object detection and semantic segmentation. Our approach leverages the inconsistency between them to identify informative samples across both tasks. We propose three constraints that specify how the tasks are coupled and introduce a method for determining the pixels belonging to the object detected by a bounding box, to later quantify the constraints as inconsistency scores. To evaluate the effectiveness of our approach, we establish multiple baselines for multi-task active learning and introduce a new metric, mean Detection Segmentation Quality (mDSQ), tailored for the multi-task active learning comparison that addresses the performance of both tasks. We conduct extensive experiments on the nuImages and A9 datasets, demonstrating that our approach outperforms existing state-of-the-art methods by up to 3.4% mDSQ on nuImages. Our approach achieves 95% of the fully-trained performance using only 67% of the available data, corresponding to 20% fewer labels compared to random selection and 5% fewer labels compared to state-of-the-art selection strategy. Our code will be made publicly available after the review process.

Ground-based solar image restoration is a computationally expensive procedure that involves nonlinear optimization techniques. The presence of atmospheric turbulence produces perturbations in individual images that make it necessary to apply blind deconvolution techniques. These techniques rely on the observation of many short exposure frames that are used to simultaneously infer the instantaneous state of the atmosphere and the unperturbed object. We have recently explored the use of machine learning to accelerate this process, with promising results. We build upon this previous work to propose several interesting improvements that lead to better models. As well, we propose a new method to accelerate the restoration based on algorithm unrolling. In this method, the image restoration problem is solved with a gradient descent method that is unrolled and accelerated aided by a few small neural networks. The role of the neural networks is to correct the estimation of the solution at each iterative step. The model is trained to perform the optimization in a small fixed number of steps with a curated dataset. Our findings demonstrate that both methods significantly reduce the restoration time compared to the standard optimization procedure. Furthermore, we showcase that these models can be trained in an unsupervised manner using observed images from three different instruments. Remarkably, they also exhibit robust generalization capabilities when applied to new datasets. To foster further research and collaboration, we openly provide the trained models, along with the corresponding training and evaluation code, as well as the training dataset, to the scientific community.

We initiate a principled study of algorithmic collective action on digital platforms that deploy machine learning algorithms. We propose a simple theoretical model of a collective interacting with a firm's learning algorithm. The collective pools the data of participating individuals and executes an algorithmic strategy by instructing participants how to modify their own data to achieve a collective goal. We investigate the consequences of this model in three fundamental learning-theoretic settings: the case of a nonparametric optimal learning algorithm, a parametric risk minimizer, and gradient-based optimization. In each setting, we come up with coordinated algorithmic strategies and characterize natural success criteria as a function of the collective's size. Complementing our theory, we conduct systematic experiments on a skill classification task involving tens of thousands of resumes from a gig platform for freelancers. Through more than two thousand model training runs of a BERT-like language model, we see a striking correspondence emerge between our empirical observations and the predictions made by our theory. Taken together, our theory and experiments broadly support the conclusion that algorithmic collectives of exceedingly small fractional size can exert significant control over a platform's learning algorithm.

Quantizing the activation, weight, and gradient to 4-bit is promising to accelerate neural network training. However, existing 4-bit training methods require custom numerical formats which are not supported by contemporary hardware. In this work, we propose a training method for transformers with all matrix multiplications implemented with the INT4 arithmetic. Training with an ultra-low INT4 precision is challenging. To achieve this, we carefully analyze the specific structures of activation and gradients in transformers to propose dedicated quantizers for them. For forward propagation, we identify the challenge of outliers and propose a Hadamard quantizer to suppress the outliers. For backpropagation, we leverage the structural sparsity of gradients by proposing bit splitting and leverage score sampling techniques to quantize gradients accurately. Our algorithm achieves competitive accuracy on a wide range of tasks including natural language understanding, machine translation, and image classification. Unlike previous 4-bit training methods, our algorithm can be implemented on the current generation of GPUs. Our prototypical linear operator implementation is up to 2.2 times faster than the FP16 counterparts and speeds up the training by up to 35.1%.

The rapid recent progress in machine learning (ML) has raised a number of scientific questions that challenge the longstanding dogma of the field. One of the most important riddles is the good empirical generalization of overparameterized models. Overparameterized models are excessively complex with respect to the size of the training dataset, which results in them perfectly fitting (i.e., interpolating) the training data, which is usually noisy. Such interpolation of noisy data is traditionally associated with detrimental overfitting, and yet a wide range of interpolating models -- from simple linear models to deep neural networks -- have recently been observed to generalize extremely well on fresh test data. Indeed, the recently discovered double descent phenomenon has revealed that highly overparameterized models often improve over the best underparameterized model in test performance. Understanding learning in this overparameterized regime requires new theory and foundational empirical studies, even for the simplest case of the linear model. The underpinnings of this understanding have been laid in very recent analyses of overparameterized linear regression and related statistical learning tasks, which resulted in precise analytic characterizations of double descent. This paper provides a succinct overview of this emerging theory of overparameterized ML (henceforth abbreviated as TOPML) that explains these recent findings through a statistical signal processing perspective. We emphasize the unique aspects that define the TOPML research area as a subfield of modern ML theory and outline interesting open questions that remain.