We develop methods, benchmarks, and training systems that turn expert data into frontier AI

The quality of training data impacts the performance of pre-trained large language models (LMs). Given a fixed budget of tokens, we study how to best select data that leads to good downstream model performance across tasks. We develop a new framework based on a simple hypothesis: just as humans acquire interdependent skills in a deliberate order, language models also follow…

Machine learning models—including prominent zero-shot models—are often trained on datasets whose labels are only a small proportion of a larger label space. Such spaces are commonly equipped with a metric that relates the labels via distances between them. We propose a simple approach to exploit this information to adapt the trained model to reliably predict new classes—or, in the case…

Recent work has shown that language models’ (LMs) prompt-based learning capabilities make them well suited for automating data labeling in domains where manual annotation is expensive. The challenge is that while writing an initial prompt is cheap, improving a prompt is costly—practitioners often require significant labeled data in order to evaluate the impact of prompt modifications. Our work asks whether…

An important class of techniques for resonant anomaly detection in high energy physics builds models that can distinguish between reference and target datasets, where only the latter has appreciable signal. Such techniques, including Classification Without Labels (CWoLa) and Simulation Assisted Likelihood-free Anomaly Detection (SALAD) rely on a single reference dataset. They cannot take advantage of commonly-available multiple datasets and thus…

Weak supervision overcomes the label bottleneck, enabling efficient development of training sets. Millions of models trained on such datasets have been deployed in the real world and interact with users on a daily basis. However, the techniques that make weak supervision attractive—such as integrating any source of signal to estimate unknown labels—also ensure that the pseudolabels it produces are highly…

Graph-based machine learning is experiencing explosive growth, driven by impressive recent developments and wide applicability. Typical approaches for graph representation learning predominantly focus on pairwise interactions, while neglecting the patterns of higher-order interactions common to complex systems. This paper explores many-body interaction models, centering on simplicial complexes. From a theoretical point of view, we offer a pair of insights illustrating…

Creating large-scale high-quality labeled datasets is a major bottleneck in supervised machine learning workflows. Auto-labeling systems are a promising way to reduce reliance on manual labeling for dataset construction. Threshold-based auto-labeling, where validation data obtained from humans is used to find a threshold for confidence above which the data is machine-labeled, is emerging as a popular solution used widely in…

The ability to generalize to unseen domains is crucial for machine learning systems, especially when we only have data from limited training domains and must deploy the resulting models in the real world. In this paper, we study domain generalization via the classic empirical risk minimization (ERM) approach with a simple regularizer based on the nuclear norm of the learned…

The challenge that climate change poses to humanity has spurred a rapidly developing field of artificial intelligence research focused on climate change applications. The climate change AI (CCAI) community works on a diverse, challenging set of problems which often involve physics-constrained ML or heterogeneous spatiotemporal data. It would be desirable to use automated machine learning (AutoML) techniques to automatically find…