Empirical studies suggest that machine learning models trained with empirical risk minimization (ERM) often rely on attributes that may be spuriously correlated with the class labels. Such models typically lead to poor performance during inference for data lacking such correlations. In this work, we explicitly consider a situation where potential spurious correlations are present in the majority of training data. In contrast with existing approaches, which use the ERM model outputs to detect the samples without spurious correlations, and either heuristically upweighting or upsampling those samples; we propose the logit correction (LC) loss, a simple yet effective improvement on the softmax cross-entropy loss, to correct the sample logit. We demonstrate that minimizing the LC loss is equivalent to maximizing the group-balanced accuracy, so the proposed LC could mitigate the negative impacts of spurious correlations. Our extensive experimental results further reveal that the proposed LC loss outperforms the SoTA solutions on multiple popular benchmarks by a large margin, an average 5.5% absolute improvement, without access to spurious attribute labels. LC is also competitive with oracle methods that make use of the attribute labels. Code is available at https://github.com/shengliu66/LC.

Large language models (LLMs) have demonstrated impressive capabilities in natural language understanding and generation, but the quality bar for medical and clinical applications is high. Today, attempts to assess models' clinical knowledge typically rely on automated evaluations on limited benchmarks. There is no standard to evaluate model predictions and reasoning across a breadth of tasks. To address this, we present MultiMedQA, a benchmark combining six existing open question answering datasets spanning professional medical exams, research, and consumer queries; and HealthSearchQA, a new free-response dataset of medical questions searched online. We propose a framework for human evaluation of model answers along multiple axes including factuality, precision, possible harm, and bias. In addition, we evaluate PaLM (a 540-billion parameter LLM) and its instruction-tuned variant, Flan-PaLM, on MultiMedQA. Using a combination of prompting strategies, Flan-PaLM achieves state-of-the-art accuracy on every MultiMedQA multiple-choice dataset (MedQA, MedMCQA, PubMedQA, MMLU clinical topics), including 67.6% accuracy on MedQA (US Medical License Exam questions), surpassing prior state-of-the-art by over 17%. However, human evaluation reveals key gaps in Flan-PaLM responses. To resolve this we introduce instruction prompt tuning, a parameter-efficient approach for aligning LLMs to new domains using a few exemplars. The resulting model, Med-PaLM, performs encouragingly, but remains inferior to clinicians. We show that comprehension, recall of knowledge, and medical reasoning improve with model scale and instruction prompt tuning, suggesting the potential utility of LLMs in medicine. Our human evaluations reveal important limitations of today's models, reinforcing the importance of both evaluation frameworks and method development in creating safe, helpful LLM models for clinical applications.

Dense prediction tasks such as segmentation and detection of pathological entities hold crucial clinical value in the digital pathology workflow. However, obtaining dense annotations on large cohorts is usually tedious and expensive. Contrastive learning (CL) is thus often employed to leverage large volumes of unlabeled data to pre-train the backbone network. To boost CL for dense prediction, some studies have proposed variations of dense matching objectives in pre-training. However, our analysis shows that employing existing dense matching strategies on histopathology images enforces invariance among incorrect pairs of dense features and, thus, is imprecise. To address this, we propose a precise location-based matching mechanism that utilizes the overlapping information between geometric transformations to precisely match regions in two augmentations. Extensive experiments on two pretraining datasets (TCGA-BRCA, NCT-CRC-HE) and three downstream datasets (GlaS, CRAG, BCSS) highlight the superiority of our method in semantic and instance segmentation tasks. Our method outperforms previous dense matching methods by up to 7.2 % in average precision for detection and 5.6 % in average precision for instance segmentation tasks. Additionally, by using our matching mechanism in the three popular contrastive learning frameworks, MoCo-v2, VICRegL and ConCL, the average precision in detection is improved by 0.7 % to 5.2 % and the average precision in segmentation is improved by 0.7 % to 4.0 %, demonstrating its generalizability.

This paper presents a novel federated reinforcement learning (Fed-RL) methodology to enhance the cyber resiliency of networked microgrids. We formulate a resilient reinforcement learning (RL) training setup which (a) generates episodic trajectories injecting adversarial actions at primary control reference signals of the grid forming (GFM) inverters and (b) trains the RL agents (or controllers) to alleviate the impact of the injected adversaries. To circumvent data-sharing issues and concerns for proprietary privacy in multi-party-owned networked grids, we bring in the aspects of federated machine learning and propose a novel Fed-RL algorithm to train the RL agents. To this end, the conventional horizontal Fed-RL approaches using decoupled independent environments fail to capture the coupled dynamics in a networked microgrid, which leads us to propose a multi-agent vertically federated variation of actor-critic algorithms, namely federated soft actor-critic (FedSAC) algorithm. We created a customized simulation setup encapsulating microgrid dynamics in the GridLAB-D/HELICS co-simulation platform compatible with the OpenAI Gym interface for training RL agents. Finally, the proposed methodology is validated with numerical examples of modified IEEE 123-bus benchmark test systems consisting of three coupled microgrids.

In molecular research, simulation \& design of molecules are key areas with significant implications for drug development, material science, and other fields. Current classical computational power falls inadequate to simulate any more than small molecules, let alone protein chains on hundreds of peptide. Therefore these experiment are done physically in wet-lab, but it takes a lot of time \& not possible to examine every molecule due to the size of the search area, tens of billions of dollars are spent every year in these research experiments. Molecule simulation \& design has lately advanced significantly by machine learning models, A fresh perspective on the issue of chemical synthesis is provided by deep generative models for graph-structured data. By optimising differentiable models that produce molecular graphs directly, it is feasible to avoid costly search techniques in the discrete and huge space of chemical structures. But these models also suffer from computational limitations when dimensions become huge and consume huge amount of resources. Quantum Generative machine learning in recent years have shown some empirical results promising significant advantages over classical counterparts.

A major direction in differentially private machine learning is differentially private fine-tuning: pretraining a model on a source of "public data" and transferring the extracted features to downstream tasks. This is an important setting because many industry deployments fine-tune publicly available feature extractors on proprietary data for downstream tasks. In this paper, we use features extracted from state-of-the-art open source models to solve benchmark tasks in computer vision and natural language processing using differentially private fine-tuning. Our key insight is that by accelerating training, we can quickly drive the model parameters to regions in parameter space where the impact of noise is minimized. In doing so, we recover the same performance as non-private fine-tuning for realistic values of epsilon in [0.01, 1.0] on benchmark image classification datasets including CIFAR100.

Contrails, short for condensation trails, are line-shaped ice clouds produced by aircraft engine exhaust when they fly through cold and humid air. They generate a greenhouse effect by absorbing or directing back to Earth approximately 33% of emitted outgoing longwave radiation. They account for over half of the climate change resulting from aviation activities. Avoiding contrails and adjusting flight routes could be an inexpensive and effective way to reduce their impact. An accurate, automated, and reliable detection algorithm is required to develop and evaluate contrail avoidance strategies. Advancement in contrail detection has been severely limited due to several factors, primarily due to a lack of quality-labeled data. Recently, proposed a large human-labeled Landsat-8 contrails dataset. Each contrail is carefully labeled with various inputs in various scenes of Landsat-8 satellite imagery. In this work, we benchmark several popular segmentation models with combinations of different loss functions and encoder backbones. This work is the first to apply state-of-the-art segmentation techniques to detect contrails in low-orbit satellite imagery. Our work can also be used as an open benchmark for contrail segmentation and is publicly available.

End-to-end multilingual ASR has become more appealing because of several reasons such as simplifying the training and deployment process and positive performance transfer from high-resource to low-resource languages. However, scaling up the number of languages, total hours, and number of unique tokens is not a trivial task. This paper explores large-scale multilingual ASR models on 70 languages. We inspect two architectures: (1) Shared embedding and output and (2) Multiple embedding and output model. In the shared model experiments, we show the importance of tokenization strategy across different languages. Later, we use our optimal tokenization strategy to train multiple embedding and output model to further improve our result. Our multilingual ASR achieves 13.9%-15.6% average WER relative improvement compared to monolingual models. We show that our multilingual ASR generalizes well on an unseen dataset and domain, achieving 9.5% and 7.5% WER on Multilingual Librispeech (MLS) with zero-shot and finetuning, respectively.

会员推理（MI）攻击突出了当前神经网络随机培训方法中的隐私弱点。然而，它为什么出现。它们仅是不完美概括的自然结果吗？在培训期间，我们应该解决哪些根本原因以减轻这些攻击？为了回答此类问题，我们提出了第一种解释MI攻击及其基于原则性因果推理的概括的方法。我们提供因果图，以定量地解释以$ 6 $攻击变体获得的观察到的MI攻击性能。我们驳斥了几种先前的非量化假设，这些假设过于简化或过度估计潜在原因的影响，从而未能捕获几个因素之间的复杂相互作用。我们的因果模型还通过共同的因果因素显示了概括和MI攻击之间的新联系。我们的因果模型具有很高的预测能力（$ 0.90 $），即它们的分析预测与经常看不见的实验中的观察结果相匹配，这使得通过它们的分析成为务实的替代方案。

提出测试释放（PTR）是一个差异隐私框架，可符合局部功能的敏感性，而不是其全球敏感性。该框架通常用于以差异性私有方式释放强大的统计数据，例如中位数或修剪平均值。尽管PTR是十年前引入的常见框架，但在诸如Robust SGD之类的应用程序中使用它，我们需要许多自适应鲁棒的查询是具有挑战性的。这主要是由于缺乏Renyi差异隐私（RDP）分析，这是一种瞬间的私人深度学习方法的基础。在这项工作中，我们概括了标准PTR，并在目标函数界定全局灵敏度时得出了第一个RDP。我们证明，与直接分析的$（\ eps，\ delta）$ -DP相比，我们的RDP绑定的PTR可以得出更严格的DP保证。我们还得出了亚采样下PTR的算法特异性隐私扩增。我们表明，我们的界限比一般的上限和接近下限的界限要紧密得多。我们的RDP界限可以为PTR的许多自适应运行的组成而更严格的隐私损失计算。作为我们的分析的应用，我们表明PTR和我们的理论结果可用于设计私人变体，用于拜占庭强大的训练算法，这些变体使用可靠的统计数据用于梯度聚集。我们对不同数据集和体系结构的标签，功能和梯度损坏的设置进行实验。我们表明，与基线相比，基于PTR的私人和强大的培训算法可显着改善该实用性。