联合学习（FL）可从分散的隐私敏感数据中学习，并在Edge客户端进行原始数据的计算。本文介绍了混合FL，其中包含在协调服务器上计算出的附加损失项（同时维护FL的私人数据限制）。有很多好处。例如，可以利用其他数据中心数据从集中式（数据中心）共同学习，并分散（联合）培训数据，并更好地匹配预期的推断数据分布。混合FL还可以将一些密集的计算（例如，将正则化）卸载到服务器中，从而大大减少了通信和客户端计算负载。对于这些和其他混合FL用例，我们提出了三种算法：平行训练，1向梯度转移和2向梯度转移。我们陈述了每种融合界限，并提供适合特定混合FL问题的直觉。最后，我们对三个任务进行了广泛的实验，表明混合FL可以将训练数据融合以达到推理分布上的准确性，并可以将通信和计算开销降低90％以上。我们的实验证实了关于算法在不同的混合FL问题设置下的性能的理论预测。

我们使用实际用户设备上的联合学习训练了一个关键字发现模型，并在部署模型以推断电话时观察到了重大改进。为了补偿在设备培训缓存中缺少的数据域，我们采用了联合联邦中心化培训。为了在没有策划标签的设备上学习，我们根据用户反馈信号制定了置信度过滤策略，用于联合蒸馏。这些技术创建了模型，可在实时A/B实验中显着改善离线评估和用户体验指标的质量指标。

通过隐私作为动机，联合学习（FL）是一种越来越多的范式，其中学习在边缘设备上集体进行，每个边缘设备都有一个用户生成的训练示例的高速缓存，该训练示例仍然驻留在本地设备上。这些在用户对其与设备的交互过程中原位收集了这些设备训练示例，因此对至少部分推断数据分布的至少一部分具有高度反射性。然而，分配换档仍然存在;设备上的培训示例可能缺少预期在推理时间遇到的一些数据输入。本文提出了一种方法来减轻这种班次：选择性使用数据中心数据，用FL混合。通过混合分散（联合）和集中（Datacenter）数据，我们可以形成有效的培训数据分布，比推断数据分布更好地匹配，导致更有用的模型，同时仍会遇到FL强加的私人培训数据访问约束。

Recent work has shown that fine-tuning large pre-trained language models on a collection of tasks described via instructions, a.k.a. instruction-tuning, improves their zero and few-shot generalization to unseen tasks. However, there is a limited understanding of the performance trade-offs of different decisions made during the instruction-tuning process. These decisions include the scale and diversity of the instruction-tuning benchmark, different task sampling strategies, fine-tuning with and without demonstrations, training using specialized datasets for reasoning and dialogue, and finally, the fine-tuning objectives themselves. In this paper, we characterize the effect of instruction-tuning decisions on downstream task performance when scaling both model and benchmark sizes. To this end, we create OPT-IML Bench: a large benchmark for Instruction Meta-Learning (IML) of 2000 NLP tasks consolidated into task categories from 8 existing benchmarks, and prepare an evaluation framework to measure three types of model generalizations: to tasks from fully held-out categories, to held-out tasks from seen categories, and to held-out instances from seen tasks. Through the lens of this framework, we first present insights about instruction-tuning decisions as applied to OPT-30B and further exploit these insights to train OPT-IML 30B and 175B, which are instruction-tuned versions of OPT. OPT-IML demonstrates all three generalization abilities at both scales on four different evaluation benchmarks with diverse tasks and input formats -- PromptSource, FLAN, Super-NaturalInstructions, and UnifiedSKG. Not only does it significantly outperform OPT on all benchmarks but is also highly competitive with existing models fine-tuned on each specific benchmark. We release OPT-IML at both scales, together with the OPT-IML Bench evaluation framework.

Traditional approaches to RL have focused on learning decision policies directly from episodic decisions, while slowly and implicitly learning the semantics of compositional representations needed for generalization. While some approaches have been adopted to refine representations via auxiliary self-supervised losses while simultaneously learning decision policies, learning compositional representations from hand-designed and context-independent self-supervised losses (multi-view) still adapts relatively slowly to the real world, which contains many non-IID subspaces requiring rapid distribution shift in both time and spatial attention patterns at varying levels of abstraction. In contrast, supervised language model cascades have shown the flexibility to adapt to many diverse manifolds, and hints of self-learning needed for autonomous task transfer. However, to date, transfer methods for language models like few-shot learning and fine-tuning still require human supervision and transfer learning using self-learning methods has been underexplored. We propose a self-supervised loss policy called contrastive distillation which manifests latent variables with high mutual information with both source and target tasks from weights to tokens. We show how this outperforms common methods of transfer learning and suggests a useful design axis of trading off compute for generalizability for online transfer. Contrastive distillation is improved through sampling from memory and suggests a simple algorithm for more efficiently sampling negative examples for contrastive losses than random sampling.

The number of international benchmarking competitions is steadily increasing in various fields of machine learning (ML) research and practice. So far, however, little is known about the common practice as well as bottlenecks faced by the community in tackling the research questions posed. To shed light on the status quo of algorithm development in the specific field of biomedical imaging analysis, we designed an international survey that was issued to all participants of challenges conducted in conjunction with the IEEE ISBI 2021 and MICCAI 2021 conferences (80 competitions in total). The survey covered participants' expertise and working environments, their chosen strategies, as well as algorithm characteristics. A median of 72% challenge participants took part in the survey. According to our results, knowledge exchange was the primary incentive (70%) for participation, while the reception of prize money played only a minor role (16%). While a median of 80 working hours was spent on method development, a large portion of participants stated that they did not have enough time for method development (32%). 25% perceived the infrastructure to be a bottleneck. Overall, 94% of all solutions were deep learning-based. Of these, 84% were based on standard architectures. 43% of the respondents reported that the data samples (e.g., images) were too large to be processed at once. This was most commonly addressed by patch-based training (69%), downsampling (37%), and solving 3D analysis tasks as a series of 2D tasks. K-fold cross-validation on the training set was performed by only 37% of the participants and only 50% of the participants performed ensembling based on multiple identical models (61%) or heterogeneous models (39%). 48% of the respondents applied postprocessing steps.

This paper describes Waymo's Collision Avoidance Testing (CAT) methodology: a scenario-based testing method that evaluates the safety of the Waymo Driver Automated Driving Systems' (ADS) intended functionality in conflict situations initiated by other road users that require urgent evasive maneuvers. Because SAE Level 4 ADS are responsible for the dynamic driving task (DDT), when engaged, without immediate human intervention, evaluating a Level 4 ADS using scenario-based testing is difficult due to the potentially infinite number of operational scenarios in which hazardous situations may unfold. To that end, in this paper we first describe the safety test objectives for the CAT methodology, including the collision and serious injury metrics and the reference behavior model representing a non-impaired eyes on conflict human driver used to form an acceptance criterion. Afterward, we introduce the process for identifying potentially hazardous situations from a combination of human data, ADS testing data, and expert knowledge about the product design and associated Operational Design Domain (ODD). The test allocation and execution strategy is presented next, which exclusively utilize simulations constructed from sensor data collected on a test track, real-world driving, or from simulated sensor data. The paper concludes with the presentation of results from applying CAT to the fully autonomous ride-hailing service that Waymo operates in San Francisco, California and Phoenix, Arizona. The iterative nature of scenario identification, combined with over ten years of experience of on-road testing, results in a scenario database that converges to a representative set of responder role scenarios for a given ODD. Using Waymo's virtual test platform, which is calibrated to data collected as part of many years of ADS development, the CAT methodology provides a robust and scalable safety evaluation.

There exists unexplained diverse variation within the predefined colon cancer stages using only features either from genomics or histopathological whole slide images as prognostic factors. Unraveling this variation will bring about improved in staging and treatment outcome, hence motivated by the advancement of Deep Neural Network libraries and different structures and factors within some genomic dataset, we aggregate atypical patterns in histopathological images with diverse carcinogenic expression from mRNA, miRNA and DNA Methylation as an integrative input source into an ensemble deep neural network for colon cancer stages classification and samples stratification into low or high risk survival groups. The results of our Ensemble Deep Convolutional Neural Network model show an improved performance in stages classification on the integrated dataset. The fused input features return Area under curve Receiver Operating Characteristic curve (AUC ROC) of 0.95 compared with AUC ROC of 0.71 and 0.68 obtained when only genomics and images features are used for the stage's classification, respectively. Also, the extracted features were used to split the patients into low or high risk survival groups. Among the 2548 fused features, 1695 features showed a statistically significant survival probability differences between the two risk groups defined by the extracted features.

Deformable image registration is a key task in medical image analysis. The Brain Tumor Sequence Registration challenge (BraTS-Reg) aims at establishing correspondences between pre-operative and follow-up scans of the same patient diagnosed with an adult brain diffuse high-grade glioma and intends to address the challenging task of registering longitudinal data with major tissue appearance changes. In this work, we proposed a two-stage cascaded network based on the Inception and TransMorph models. The dataset for each patient was comprised of a native pre-contrast (T1), a contrast-enhanced T1-weighted (T1-CE), a T2-weighted (T2), and a Fluid Attenuated Inversion Recovery (FLAIR). The Inception model was used to fuse the 4 image modalities together and extract the most relevant information. Then, a variant of the TransMorph architecture was adapted to generate the displacement fields. The Loss function was composed of a standard image similarity measure, a diffusion regularizer, and an edge-map similarity measure added to overcome intensity dependence and reinforce correct boundary deformation. We observed that the addition of the Inception module substantially increased the performance of the network. Additionally, performing an initial affine registration before training the model showed improved accuracy in the landmark error measurements between pre and post-operative MRIs. We observed that our best model composed of the Inception and TransMorph architectures while using an initially affine registered dataset had the best performance with a median absolute error of 2.91 (initial error = 7.8). We achieved 6th place at the time of model submission in the final testing phase of the BraTS-Reg challenge.

Motion planning and control in autonomous car racing are one of the most challenging and safety-critical tasks due to high speed and dynamism. The lower-level control nodes are expected to be highly optimized due to resource constraints of onboard embedded processing units, although there are strict latency requirements. Some of these guarantees can be provided at the application level, such as using ROS2's Real-Time executors. However, the performance can be far from satisfactory as many modern control algorithms (such as Model Predictive Control) rely on solving complicated online optimization problems at each iteration. In this paper, we present a simple yet effective multi-threading technique to optimize the throughput of online-control algorithms for resource-constrained autonomous racing platforms. We achieve this by maintaining a systematic pool of worker threads solving the optimization problem in parallel which can improve the system performance by reducing latency between control input commands. We further demonstrate the effectiveness of our method using the Model Predictive Contouring Control (MPCC) algorithm running on Nvidia's Xavier AGX platform.