Machine learning systems deployed in the wild are often trained on a source distribution but deployed on a different target distribution. Unlabeled data can be a powerful point of leverage for mitigating these distribution shifts, as it is frequently much more available than labeled data. However, existing distribution shift benchmarks for unlabeled data do not reflect the breadth of scenarios that arise in real-world applications. In this work, we present the WILDS 2.0 update, which extends 8 of the 10 datasets in the WILDS benchmark of distribution shifts to include curated unlabeled data that would be realistically obtainable in deployment. To maintain consistency, the labeled training, validation, and test sets, as well as the evaluation metrics, are exactly the same as in the original WILDS benchmark. These datasets span a wide range of applications (from histology to wildlife conservation), tasks (classification, regression, and detection), and modalities (photos, satellite images, microscope slides, text, molecular graphs). We systematically benchmark state-of-the-art methods that leverage unlabeled data, including domain-invariant, self-training, and self-supervised methods, and show that their success on WILDS 2.0 is limited. To facilitate method development and evaluation, we provide an open-source package that automates data loading and contains all of the model architectures and methods used in this paper. Code and leaderboards are available at https://wilds.stanford.edu.
Training a model for grammatical error correction (GEC) requires a set of labeled ungrammatical / grammatical sentence pairs, but manually annotating such pairs can be expensive. Recently, the Break-It-Fix-It (BIFI) framework has demonstrated strong results on learning to repair a broken program without any labeled examples, but this relies on a perfect critic (e.g., a compiler) that returns whether an example is valid or not, which does not exist for the GEC task. In this work, we show how to leverage a pretrained language model (LM) in defining an LM-Critic, which judges a sentence to be grammatical if the LM assigns it a higher probability than its local perturbations. We apply this LM-Critic and BIFI along with a large set of unlabeled sentences to bootstrap realistic ungrammatical / grammatical pairs for training a corrector. We evaluate our approach on GEC datasets across multiple domains (CoNLL-2014, BEA-2019, GMEG-wiki and GMEG-yahoo) and show that it outperforms existing methods in both the unsupervised setting (+7.7 F0.5) and the supervised setting (+0.5 F0.5).
AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and are adaptable to a wide range of downstream tasks. We call these models foundation models to underscore their critically central yet incomplete character. This report provides a thorough account of the opportunities and risks of foundation models, ranging from their capabilities (e.g., language, vision, robotics, reasoning, human interaction) and technical principles(e.g., model architectures, training procedures, data, systems, security, evaluation, theory) to their applications (e.g., law, healthcare, education) and societal impact (e.g., inequity, misuse, economic and environmental impact, legal and ethical considerations). Though foundation models are based on standard deep learning and transfer learning, their scale results in new emergent capabilities,and their effectiveness across so many tasks incentivizes homogenization. Homogenization provides powerful leverage but demands caution, as the defects of the foundation model are inherited by all the adapted models downstream. Despite the impending widespread deployment of foundation models, we currently lack a clear understanding of how they work, when they fail, and what they are even capable of due to their emergent properties. To tackle these questions, we believe much of the critical research on foundation models will require deep interdisciplinary collaboration commensurate with their fundamentally sociotechnical nature.
We consider repair tasks: given a critic (e.g., compiler) that assesses the quality of an input, the goal is to train a fixer that converts a bad example (e.g., code with syntax errors) into a good one (e.g., code with no syntax errors). Existing works create training data consisting of (bad, good) pairs by corrupting good examples using heuristics (e.g., dropping tokens). However, fixers trained on this synthetically-generated data do not extrapolate well to the real distribution of bad inputs. To bridge this gap, we propose a new training approach, Break-It-Fix-It (BIFI), which has two key ideas: (i) we use the critic to check a fixer's output on real bad inputs and add good (fixed) outputs to the training data, and (ii) we train a breaker to generate realistic bad code from good code. Based on these ideas, we iteratively update the breaker and the fixer while using them in conjunction to generate more paired data. We evaluate BIFI on two code repair datasets: GitHub-Python, a new dataset we introduce where the goal is to repair Python code with AST parse errors; and DeepFix, where the goal is to repair C code with compiler errors. BIFI outperforms existing methods, obtaining 90.5% repair accuracy on GitHub-Python (+28.5%) and 71.7% on DeepFix (+5.6%). Notably, BIFI does not require any labeled data; we hope it will be a strong starting point for unsupervised learning of various repair tasks.
The problem of answering questions using knowledge from pre-trained language models (LMs) and knowledge graphs (KGs) presents two challenges: given a QA context (question and answer choice), methods need to (i) identify relevant knowledge from large KGs, and (ii) perform joint reasoning over the QA context and KG. Here we propose a new model, QA-GNN, which addresses the above challenges through two key innovations: (i) relevance scoring, where we use LMs to estimate the importance of KG nodes relative to the given QA context, and (ii) joint reasoning, where we connect the QA context and KG to form a joint graph, and mutually update their representations through graph-based message passing. We evaluate QA-GNN on the CommonsenseQA and OpenBookQA datasets, and show its improvement over existing LM and LM+KG models, as well as its capability to perform interpretable and structured reasoning, e.g., correctly handling negation in questions.
Distribution shifts can cause significant degradation in a broad range of machine learning (ML) systems deployed in the wild. However, many widely-used datasets in the ML community today were not designed for evaluating distribution shifts. These datasets typically have training and test sets drawn from the same distribution, and prior work on retrofitting them with distribution shifts has generally relied on artificial shifts that need not represent the kinds of shifts encountered in the wild. In this paper, we present WILDS, a benchmark of in-the-wild distribution shifts spanning diverse data modalities and applications, from tumor identification to wildlife monitoring to poverty mapping. WILDS builds on top of recent data collection efforts by domain experts in these applications and provides a unified collection of datasets with evaluation metrics and train/test splits that are representative of real-world distribution shifts. These datasets reflect distribution shifts arising from training and testing on different hospitals, cameras, countries, time periods, demographics, molecular scaffolds, etc., all of which cause substantial performance drops in our baseline models. Finally, we survey other applications that would be promising additions to the benchmark but for which we did not manage to find appropriate datasets; we discuss their associated challenges and detail datasets and shifts where we did not see an appreciable performance drop. By unifying datasets from a variety of application areas and making them accessible to the ML community, we hope to encourage the development of general-purpose methods that are anchored to real-world distribution shifts and that work well across different applications and problem settings. Data loaders, default models, and leaderboards are available at https://wilds.stanford.edu.
We consider the problem of learning to repair programs from diagnostic feedback (e.g., compiler error messages). Program repair is challenging for two reasons: First, it requires reasoning and tracking symbols across source code and diagnostic feedback. Second, labeled datasets available for program repair are relatively small. In this work, we propose novel solutions to these two challenges. First, we introduce a program-feedback graph, which connects symbols relevant to program repair in source code and diagnostic feedback, and then apply a graph neural network on top to model the reasoning process. Second, we present a self-supervised learning paradigm for program repair that leverages unlabeled programs available online to create a large amount of extra program repair examples, which we use to pre-train our models. We evaluate our proposed approach on two applications: correcting introductory programming assignments (DeepFix dataset) and correcting the outputs of program synthesis (SPoC dataset). Our final system, DrRepair, significantly outperforms prior work, achieving 66.1% full repair rate on DeepFix (+20.8% over the prior best), and 48.0% synthesis success rate on SPoC (+3.3% over the prior best).
Automated analysis of clinical notes is attracting increasing attention. However, there has not been much work on medical term abbreviation disambiguation. Such abbreviations are abundant, and highly ambiguous, in clinical documents. One of the main obstacles is the lack of large scale, balance labeled data sets. To address the issue, we propose a few-shot learning approach to take advantage of limited labeled data. Specifically, a neural topic-attention model is applied to learn improved contextualized sentence representations for medical term abbreviation disambiguation. Another vital issue is that the existing scarce annotations are noisy and missing. We re-examine and correct an existing dataset for training and collect a test set to evaluate the models fairly especially for rare senses. We train our model on the training set which contains 30 abbreviation terms as categories (on average, 479 samples and 3.24 classes in each term) selected from a public abbreviation disambiguation dataset, and then test on a manually-created balanced dataset (each class in each term has 15 samples). We show that enhancing the sentence representation with topic information improves the performance on small-scale unbalanced training datasets by a large margin, compared to a number of baseline models.