One of today's goals for industrial robot systems is to allow fast and easy provisioning for new tasks. Skill-based systems that use planning and knowledge representation have long been one possible answer to this. However, especially with contact-rich robot tasks that need careful parameter settings, such reasoning techniques can fall short if the required knowledge not adequately modeled. We show an approach that provides a combination of task-level planning and reasoning with targeted learning of skill parameters for a task at hand. Starting from a task goal formulated in PDDL, the learnable parameters in the plan are identified and an operator can choose reward functions and parameters for the learning process. A tight integration with a knowledge framework allows to form a prior for learning and the usage of multi-objective Bayesian optimization eases to balance aspects such as safety and task performance that can often affect each other. We demonstrate the efficacy and versatility of our approach by learning skill parameters for two different contact-rich tasks and show their successful execution on a real 7-DOF KUKA-iiwa.

机器人技能系统旨在减少机器人设置时间的新制造任务。但是，对于灵巧，接触术的任务，通常很难找到正确的技能参数。一种策略是通过允许机器人系统直接学习任务来学习这些参数。对于学习问题，机器人操作员通常可以指定参数值的类型和范围。然而，鉴于他们先前的经验，机器人操作员应该能够通过提供有关在参数空间中找到最佳解决方案的知识猜测，从而进一步帮助学习过程。有趣的是，当前的机器人学习框架中没有利用这种先验知识。我们介绍了一种结合用户先验和贝叶斯优化的方法，以便在机器人部署时间快速优化机器人工业任务。我们在模拟中学习的三个任务以及直接在真实机器人系统上学习的两个任务中学习了我们的方法。此外，我们通过自动从良好表现的配置中自动构造先验来从相应的仿真任务中转移知识，以在真实系统上学习。为了处理潜在的任务目标，任务被建模为多目标问题。我们的结果表明，操作员的先验是用户指定和转移的，大大加快了富丽堂皇的阵线的发现，并且通常产生的最终性能远远超过了拟议的基线。

增强学习（RL）是一个强大的数学框架，可让机器人通过反复试验学习复杂的技能。尽管在许多应用中取得了许多成功，但RL算法仍然需要数千个试验才能融合到高性能的政策，可以在学习时产生危险的行为，并且优化的政策（通常为神经网络建模）几乎可以在无法执行的解释时给出零的解释。任务。由于这些原因，在工业环境中采用RL并不常见。另一方面，行为树（BTS）可以提供一个策略表示，a）支持模块化和可综合的技能，b）允许轻松解释机器人动作，c）提供了有利的低维参数空间。在本文中，我们提出了一种新颖的算法，该算法可以学习模拟中BT策略的参数，然后在没有任何其他培训的情况下将其推广到物理机器人。我们利用了使用数字化工作站的物理模拟器，并使用黑盒优化器优化相关参数。我们在包括避免障碍物和富含接触的插入（孔洞）的任务中，通过7道型kuka-iiwa操纵器展示了我们方法的功效，其中我们的方法优于基准。

A major goal of multimodal research is to improve machine understanding of images and text. Tasks include image captioning, text-to-image generation, and vision-language representation learning. So far, research has focused on the relationships between images and text. For example, captioning models attempt to understand the semantics of images which are then transformed into text. An important question is: which annotation reflects best a deep understanding of image content? Similarly, given a text, what is the best image that can present the semantics of the text? In this work, we argue that the best text or caption for a given image is the text which would generate the image which is the most similar to that image. Likewise, the best image for a given text is the image that results in the caption which is best aligned with the original text. To this end, we propose a unified framework that includes both a text-to-image generative model and an image-to-text generative model. Extensive experiments validate our approach.

Variance parameter estimation in linear mixed models is a challenge for many classical nonlinear optimization algorithms due to the positive-definiteness constraint of the random effects covariance matrix. We take a completely novel view on parameter estimation in linear mixed models by exploiting the intrinsic geometry of the parameter space. We formulate the problem of residual maximum likelihood estimation as an optimization problem on a Riemannian manifold. Based on the introduced formulation, we give geometric higher-order information on the problem via the Riemannian gradient and the Riemannian Hessian. Based on that, we test our approach with Riemannian optimization algorithms numerically. Our approach yields a higher quality of the variance parameter estimates compared to existing approaches.

Machine-learning models are increasingly used to predict properties of atoms in chemical systems. There have been major advances in developing descriptors and regression frameworks for this task, typically starting from (relatively) small sets of quantum-mechanical reference data. Larger datasets of this kind are becoming available, but remain expensive to generate. Here we demonstrate the use of a large dataset that we have "synthetically" labelled with per-atom energies from an existing ML potential model. The cheapness of this process, compared to the quantum-mechanical ground truth, allows us to generate millions of datapoints, in turn enabling rapid experimentation with atomistic ML models from the small- to the large-data regime. This approach allows us here to compare regression frameworks in depth, and to explore visualisation based on learned representations. We also show that learning synthetic data labels can be a useful pre-training task for subsequent fine-tuning on small datasets. In the future, we expect that our open-sourced dataset, and similar ones, will be useful in rapidly exploring deep-learning models in the limit of abundant chemical data.

The last few years have seen a lot of work to address the challenge of low-latency and high-throughput convolutional neural network inference. Integrated photonics has the potential to dramatically accelerate neural networks because of its low-latency nature. Combined with the concept of Joint Transform Correlator (JTC), the computationally expensive convolution functions can be computed instantaneously (time of flight of light) with almost no cost. This 'free' convolution computation provides the theoretical basis of the proposed PhotoFourier JTC-based CNN accelerator. PhotoFourier addresses a myriad of challenges posed by on-chip photonic computing in the Fourier domain including 1D lenses and high-cost optoelectronic conversions. The proposed PhotoFourier accelerator achieves more than 28X better energy-delay product compared to state-of-art photonic neural network accelerators.

We present a smoothly broken power law functional form that accurately models and extrapolates the scaling behaviors of deep neural networks (i.e. how the evaluation metric of interest varies as the amount of compute used for training, number of model parameters, training dataset size, or upstream performance varies) for each task within a large and diverse set of upstream and downstream tasks, in zero-shot, prompted, and fine-tuned settings. This set includes large-scale vision and unsupervised language tasks, diffusion generative modeling of images, arithmetic, and reinforcement learning. When compared to other functional forms for neural scaling behavior, this functional form yields extrapolations of scaling behavior that are considerably more accurate on this set. Moreover, this functional form accurately models and extrapolates scaling behavior that other functional forms are incapable of expressing such as the non-monotonic transitions present in the scaling behavior of phenomena such as double descent and the delayed, sharp inflection points present in the scaling behavior of tasks such as arithmetic. Lastly, we use this functional form to glean insights about the limit of the predictability of scaling behavior. Code is available at https://github.com/ethancaballero/broken_neural_scaling_laws

我们提供了奖励黑客的第一个正式定义，即优化不完美的代理奖励功能的现象，$ \ Mathcal {\ tilde {r}} $，根据真实的奖励功能，$ \ MATHCAL {R} $导致性能差。 。我们说，如果增加预期的代理回报率永远无法减少预期的真实回报，则代理是不可接受的。直觉上，可以通过从奖励功能（使其“较窄”）中留出一些术语或忽略大致等效的结果之间的细粒度区分来创建一个不可接受的代理，但是我们表明情况通常不是这样。一个关键的见解是，奖励的线性性（在州行动访问计数中）使得无法实现的状况非常强烈。特别是，对于所有随机策略的集合，只有在其中一个是恒定的，只有两个奖励函数才能是不可接受的。因此，我们将注意力转移到确定性的政策和有限的随机政策集中，在这些策略中，始终存在非平凡的不可动摇的对，并为简化的存在建立必要和充分的条件，这是一个重要的不被限制的特殊情况。我们的结果揭示了使用奖励函数指定狭窄任务和对齐人类价值的AI系统之间的紧张关系。

现代机器学习研究依赖于相对较少的精心策划数据集。即使在这些数据集中，通常在“不整合”或原始数据中，从业人员也面临着重要的数据质量和多样性问题，这些问题可能会非常强烈地解决。应对这些挑战的现有方法往往会对特定问题做出强烈的假设，并且通常需要先验知识或元数据，例如域标签。我们的工作与这些方法是正交的：相反，我们专注于为元数据考古学提供一个统一和有效的框架 - 在数据集中发现和推断示例的元数据。我们使用简单的转换策划了可能存在的数据集（例如，错误标记，非典型或过度分布示例）中可能存在的数据子集，并利用这些探针套件之间的学习动力学差异来推断感兴趣的元数据。我们的方法与跨不同任务的更复杂的缓解方法相提并论：识别和纠正标签错误的示例，对少数民族样本进行分类，优先考虑与培训相关的点并启用相关示例的可扩展人类审核。