Models trained on a labeled source domain often generalize poorly when deployed on an out-of-distribution (OOD) target domain. In the domain adaptation setting where unlabeled target data is available, self-supervised pretraining (e.g., contrastive learning or masked autoencoding) is a promising method to mitigate this performance drop. Pretraining depends on generic data augmentations (e.g., cropping or masking) to learn representations that generalize across domains, which may not work for all distribution shifts. In this paper, we show on real-world tasks that standard fine-tuning after pretraining does not consistently improve OOD error over simply training from scratch on labeled source data. To better leverage pretraining for distribution shifts, we propose the Connect Later framework, which fine-tunes the model with targeted augmentations designed with knowledge of the shift. Intuitively, pretraining learns good representations within the source and target domains, while fine-tuning with targeted augmentations improves generalization across domains. Connect Later achieves state-of-the-art OOD accuracy while maintaining comparable or better in-distribution accuracy on 4 real-world tasks in wildlife identification (iWildCam-WILDS), tumor detection (Camelyon17-WILDS), and astronomy (AstroClassification, Redshifts).

Language modeling on large-scale datasets improves performance of various downstream tasks. The validation pre-training loss is often used as the evaluation metric for language models since the pre-training loss tends to be well-correlated with downstream performance (which is itself hard to evaluate comprehensively). Contrary to the conventional wisdom, this paper shows that 1) pre-training loss cannot fully explain downstream performance and 2) flatness of the model is well-correlated with downstream performance where pre-training loss is not. We identify three ways to produce models with the same pre-training loss but different downstream performance: continue pre-training after convergence, increasing the model size, and changing the pre-training algorithms. These experiments demonstrate the existence of implicit bias of pre-training algorithms---among models with the same minimal pre-training loss, they implicitly prefer more transferable ones. Toward understanding this implicit bias, we prove that SGD with standard mini-batch noise implicitly prefers flatter minima of pre-training loss in language models, and empirically observe a strong correlation between flatness (measured by the trace of Hessian) and downstream performance among models with the same pre-training loss. We also prove in a synthetic language setting that among models with the minimal pre-training loss, the flattest model transfers to downstream tasks.

The mixture proportions of pretraining data domains (e.g., Wikipedia, books, web text) greatly affect language model (LM) performance. In this paper, we propose Domain Reweighting with Minimax Optimization (DoReMi), which first trains a small proxy model using group distributionally robust optimization (Group DRO) over domains to produce domain weights (mixture proportions) without knowledge of downstream tasks. We then resample a dataset with these domain weights and train a larger, full-sized model. In our experiments, we use DoReMi on a 280M-parameter proxy model to set the domain weights for training an 8B-parameter model (30x larger) more efficiently. On The Pile, DoReMi improves perplexity across all domains, even when it downweights a domain. DoReMi improves average few-shot downstream accuracy by 6.5% points over a baseline model trained using The Pile's default domain weights and reaches the baseline accuracy with 2.6x fewer training steps. On the GLaM dataset, DoReMi, which has no knowledge of downstream tasks, even matches the performance of using domain weights tuned on downstream tasks.

Selecting a suitable pretraining dataset is crucial for both general-domain (e.g., GPT-3) and domain-specific (e.g., Codex) language models (LMs). We formalize this problem as selecting a subset of a large raw unlabeled dataset to match a desired target distribution given unlabeled target samples. Due to the scale and dimensionality of the raw text data, existing methods use simple heuristics or require human experts to manually curate data. Instead, we extend the classic importance resampling approach used in low-dimensions for LM data selection. We propose Data Selection with Importance Resampling (DSIR), an efficient and scalable framework that estimates importance weights in a reduced feature space for tractability and selects data with importance resampling according to these weights. We instantiate the DSIR framework with hashed n-gram features for efficiency, enabling the selection of 100M documents from the full Pile dataset in 4.5 hours. To measure whether hashed n-gram features preserve the aspects of the data that are relevant to the target, we define KL reduction, a data metric that measures the proximity between the selected pretraining data and the target on some feature space. Across 8 data selection methods (including expert selection), KL reduction on hashed n-gram features highly correlates with average downstream accuracy (r=0.82). When selecting data for continued pretraining on a specific domain, DSIR performs comparably to expert curation across 8 target distributions. When pretraining general-domain models (target is Wikipedia and books), DSIR improves over random selection and heuristic filtering baselines by 2--2.5% on the GLUE benchmark.

We consider unsupervised domain adaptation (UDA), where labeled data from a source domain (e.g., photos) and unlabeled data from a target domain (e.g., sketches) are used to learn a classifier for the target domain. Conventional UDA methods (e.g., domain adversarial training) learn domain-invariant features to generalize from the source domain to the target domain. In this paper, we show that contrastive pre-training, which learns features on unlabeled source and target data and then fine-tunes on labeled source data, is competitive with strong UDA methods. However, we find that contrastive pre-training does not learn domain-invariant features, diverging from conventional UDA intuitions. We show theoretically that contrastive pre-training can learn features that vary subtantially across domains but still generalize to the target domain, by disentangling domain and class information. We empirically validate our theory on benchmark vision datasets.

Distribution shifts—where the training distribution differs from the test distribution—can substantially degrade the accuracy of machine learning (ML) systems deployed in the wild. Despite their ubiquity in the real-world deployments, these distribution shifts are under-represented in the datasets widely used in the ML community today. To address this gap, we present WILDS, a curated benchmark of 10 datasets reflecting a diverse range of distribution shifts that naturally arise in real-world applications, such as shifts across hospitals for tumor identification; across camera traps for wildlife monitoring; and across time and location in satellite imaging and poverty mapping. On each dataset, we show that standard training yields substantially lower out-of-distribution than in-distribution performance. This gap remains even with models trained by existing methods for tackling distribution shifts, underscoring the need for new methods for training models that are more robust to the types of distribution shifts that arise in practice. To facilitate method development, we provide an open-source package that automates dataset loading, contains default model architectures and hyperparameters, and standardizes evaluations. The full paper, code, and leaderboards are available at https://wilds.stanford.edu.

We focus on prediction problems with structured outputs that are subject to output validity constraints, e.g. pseudocode-to-code translation where the code must compile. While labeled input-output pairs are expensive to obtain, "unlabeled" outputs, i.e. outputs without corresponding inputs, are freely available (e.g. code on GitHub) and provide information about output validity. Pre-training captures this structure by training a denoiser to denoise corrupted versions of unlabeled outputs. We first show that standard fine-tuning after pre-training destroys some of this structure. We then propose composed fine-tuning, which trains a predictor composed with the pre-trained denoiser. Importantly, the denoiser is fixed to preserve output structure. Like standard fine-tuning, the predictor is also initialized with the pre-trained denoiser. We prove for two-layer ReLU networks that composed fine-tuning significantly reduces the complexity of the predictor, thus improving generalization. Empirically, we show that composed fine-tuning improves over standard fine-tuning on two pseudocode-to-code translation datasets (3% and 6% relative). The improvement is magnified on out-of-distribution (OOD) examples (4% and 25% relative), suggesting that reducing predictor complexity improves OOD extrapolation.

Pretrained language models have achieved state-of-the-art performance when adapted to a downstream NLP task. However, theoretical analysis of these models is scarce and challenging since the pretraining and downstream tasks can be very different. We propose an analysis framework that links the pretraining and downstream tasks with an underlying latent variable generative model of text -- the downstream classifier must recover a function of the posterior distribution over the latent variables. We analyze head tuning (learning a classifier on top of the frozen pretrained model) and prompt tuning in this setting. The generative model in our analysis is either a Hidden Markov Model (HMM) or an HMM augmented with a latent memory component, motivated by long-term dependencies in natural language. We show that 1) under certain non-degeneracy conditions on the HMM, simple classification heads can solve the downstream task, 2) prompt tuning obtains downstream guarantees with weaker non-degeneracy conditions, and 3) our recovery guarantees for the memory-augmented HMM are stronger than for the vanilla HMM because task-relevant information is easier to recover from the long-term memory. Experiments on synthetically generated data from HMMs back our theoretical findings.

Adversarial training augments the training set with perturbations to improve the robust error (over worst-case perturbations), but it often leads to an increase in the standard error (on unperturbed test inputs). Previous explanations for this tradeoff rely on the assumption that no predictor in the hypothesis class has low standard and robust error. In this work, we precisely characterize the effect of augmentation on the standard error in linear regression when the optimal linear predictor has zero standard and robust error. In particular, we show that the standard error could increase even when the augmented perturbations have noiseless observations from the optimal linear predictor. We then prove that the recently proposed robust self-training (RST) estimator improves robust error without sacrificing standard error for noiseless linear regression. Empirically, for neural networks, we find that RST with different adversarial training methods improves both standard and robust error for random and adversarial rotations and adversarial l_infty perturbations in CIFAR-10.

Many machine learning tasks require sampling a subset of items from a collection based on a parameterized distribution. The Gumbel-softmax trick can be used to sample a single item, and allows for low-variance reparameterized gradients with respect to the parameters of the underlying distribution. However, stochastic optimization involving subset sampling is typically not reparameterizable. To overcome this limitation, we define a continuous relaxation of subset sampling that provides reparameterization gradients by generalizing the Gumbel-max trick. We use this approach to sample subsets of features in an instance-wise feature selection task for model interpretability, subsets of neighbors to implement a deep stochastic k-nearest neighbors model, and sub-sequences of neighbors to implement parametric t-SNE by directly comparing the identities of local neighbors. We improve performance in all these tasks by incorporating subset sampling in end-to-end training.

Large amounts of labeled data are typically required to train deep learning models. For many real-world problems, however, acquiring additional data can be expensive or even impossible. We present semi-supervised deep kernel learning (SSDKL), a semi-supervised regression model based on minimizing predictive variance in the posterior regularization framework. SSDKL combines the hierarchical representation learning of neural networks with the probabilistic modeling capabilities of Gaussian processes. By leveraging unlabeled data, we show improvements on a diverse set of real-world regression tasks over supervised deep kernel learning and semi-supervised methods such as VAT and mean teacher adapted for regression.

The lack of reliable data in developing countries is a major obstacle to sustainable development, food security, and disaster relief. Poverty data, for example, is typically scarce, sparse in coverage, and labor-intensive to obtain. Remote sensing data such as high-resolution satellite imagery, on the other hand, is becoming increasingly available and inexpensive. Unfortunately, such data is highly unstructured and currently no techniques exist to automatically extract useful insights to inform policy decisions and help direct humanitarian efforts. We propose a novel machine learning approach to extract large-scale socioeconomic indicators from high-resolution satellite imagery. The main challenge is that training data is very scarce, making it difficult to apply modern techniques such as Convolutional Neural Networks (CNN). We therefore propose a transfer learning approach where nighttime light intensities are used as a data-rich proxy. We train a fully convolutional CNN model to predict nighttime lights from daytime imagery, simultaneously learning features that are useful for poverty prediction. The model learns filters identifying different terrains and man-made structures, including roads, buildings, and farmlands, without any supervision beyond nighttime lights. We demonstrate that these learned features are highly informative for poverty mapping, even approaching the predictive performance of survey data collected in the field.