Active testing enables label-efficient evaluation of predictive models through careful data acquisition, but it can pose a significant computational cost. We identify cost-saving measures that enable active testing to be scaled up to large language models (LLMs). In particular we show that the surrogate model used to guide data acquisition can be constructed cheaply using in-context learning, does not require updating within an active-testing loop, and can be smaller than the target model. We even find we can make good data-acquisition decisions without making predictions with the target model. As a result we are able to achieve much more accurate evaluations of LLM performance relative to using randomly acquired data. We additionally introduce a bootstrap estimator of evaluation error, which we show to be a useful indicator of how well active testing is working within a single run.

Active testing enables label-efficient evaluation of predictive models through careful data acquisition, but it can pose a significant computational cost. We identify cost-saving measures that enable active testing to be scaled up to large language models (LLMs). In particular we show that the surrogate model used to guide data acquisition can be constructed cheaply using in-context learning, does not require updating within an active-testing loop, and can be smaller than the target model. We even find we can make good data-acquisition decisions without making predictions with the target model. As a result we are able to achieve much more accurate evaluations of LLM performance relative to using randomly acquired data. We additionally introduce a bootstrap estimator of evaluation error, which we show to be a useful indicator of how well active testing is working within a single run.

In active learning (AL), we focus on reducing the data annotation cost from the model training perspective. However, testing'', which often refers to the model evaluation process of using empirical risk to estimate the intractable true generalization risk, also requires data annotations. The annotation cost for testing'' (model evaluation) is under-explored. Even in works that study active model evaluation or active testing (AT), the learning and testing ends are disconnected. In this paper, we propose a novel active testing while learning (ATL) framework that integrates active learning with active testing.

In Figure 2, we examine the probability of acquiring a '7' as a function of the number of acquired We see that XWED initially focuses on 7s but then diversifies. The XWED behavior is preferable: we are initially unsure about the loss of these points, but once the loss is well characterized for the 7s we should explore other areas as well. B.2 Constant π Fails for Distribution Shift. Figure B.1 (a) shows that, for LURE suffered high variance in Figure 3. In Figure B.1 (b), we observe that ASE continues to Figure B.2 demonstrates that ASEs continue to outperform all other baselines for the task of This result highlights the importance of the adaptive nature of both ASE-and LUREbased active testing. Figure B.2: V ariant of the experiments of 7.3 where we estimate the accuracy of the main model. We here investigate a variation of the experiments in 7.3: reducing the size of the training set to Despite this, Figure B.3 demonstrates that ASEs continue to outperform all baselines.

Active testing enables label-efficient evaluation of models through careful data acquisition. However, its significant computational costs have previously undermined its use for large models. We show how it can be successfully scaled up to the evaluation of large language models (LLMs). In particular we show that the surrogate model used to guide data acquisition can be constructed cheaply using in-context learning, does not require updating within an active-testing loop, and can be smaller than the target model. We even find we can make good data-acquisition decisions without computing predictions with the target model and further introduce a single-run error estimator to asses how well active testing is working on the fly. We find that our approach is able to more effectively evaluate LLM performance with less data than current standard practices.

In active learning (AL), we focus on reducing the data annotation cost from the model training perspective. However, "testing'', which often refers to the model evaluation process of using empirical risk to estimate the intractable true generalization risk, also requires data annotations. The annotation cost for "testing'' (model evaluation) is under-explored. Even in works that study active model evaluation or active testing (AT), the learning and testing ends are disconnected. In this paper, we propose a novel active testing while learning (ATL) framework that integrates active learning with active testing.

We introduce active testing: a new framework for sample-efficient model evaluation. While approaches like active learning reduce the number of labels needed for model training, existing literature largely ignores the cost of labeling test data, typically unrealistically assuming large test sets for model evaluation. This creates a disconnect to real applications where test labels are important and just as expensive, e.g. for optimizing hyperparameters. Active testing addresses this by carefully selecting the test points to label, ensuring model evaluation is sample-efficient. To this end, we derive theoretically-grounded and intuitive acquisition strategies that are specifically tailored to the goals of active testing, noting these are distinct to those of active learning. Actively selecting labels introduces a bias; we show how to remove that bias while reducing the variance of the estimator at the same time. Active testing is easy to implement, effective, and can be applied to any supervised machine learning method. We demonstrate this on models including WideResNet and Gaussian processes on datasets including CIFAR-100.