Related papers: If It's Nice, Do It Twice: We Should Try Iterative…

The discovery and optimization of materials for specific applications is hampered by the practically infinite number of possible elemental combinations and associated properties, also known as the `combinatorial explosion'. By nature of the…

Foundation models are trained on increasingly immense and opaque datasets. Even while these models are now key in AI system building, it can be difficult to answer the straightforward question: has the model already encountered a given…

Iterative self-improvement fine-tunes an autoregressive large language model (LLM) on reward-verified outputs generated by the LLM itself. In contrast to the empirical success of self-improvement, the theoretical foundation of this…

Machine Learning models increasingly face data integrity challenges due to the use of large-scale training datasets drawn from the Internet. We study what model developers can do if they detect that some data was manipulated or incorrect.…

Untrusted data used to train a model might have been manipulated to endow the learned model with hidden properties that the data contributor might later exploit. Data purification aims to remove such manipulations prior to training the…

As demand for large corpora increases with the size of current state-of-the-art language models, using web data as the main part of the pre-training corpus for these models has become a ubiquitous practice. This, in turn, has introduced an…

Recent research shows that fine-tuning on benign instruction-following data can inadvertently undo the safety alignment process and increase a model's propensity to comply with harmful queries. While instruction-following fine-tuning is…

The public text record -- the material from which both people and AI systems now learn -- is increasingly shaped by its own outputs. Generated text enters the public record, later agents learn from it, and the cycle repeats. Here we develop…

Iterative self-training (self-distillation) repeatedly refits a model on pseudo-labels generated by its own predictions. We study this procedure in overparameterized linear regression: an initial estimator is trained on noisy labels, and…

Recent advances in generative models have made it increasingly difficult to distinguish real data from model-generated synthetic data. Using synthetic data for successive training of future model generations creates "self-consuming loops",…

As Machine Learning models continue to be relied upon for making automated decisions, the issue of model bias becomes more and more prevalent. In this paper, we approach training a text classifica-tion model and optimize on bias…

Deep generative models have made tremendous progress in modeling complex data, often exhibiting generation quality that surpasses a typical human's ability to discern the authenticity of samples. Undeniably, a key driver of this success is…

Most previous work on the recently developed language-modeling approach to information retrieval focuses on document-specific characteristics, and therefore does not take into account the structure of the surrounding corpus. We propose a…

Collaborative learning has gained great popularity due to its benefit of data privacy protection: participants can jointly train a Deep Learning model without sharing their training sets. However, recent works discovered that an adversary…

Language models (LMs) can reproduce (or amplify) toxic language seen during training, which poses a risk to their practical application. In this paper, we conduct extensive experiments to study this phenomenon. We analyze the impact of…

Multilinguality is a core capability for modern foundation models, yet training high-quality multilingual models remains challenging due to uneven data availability across languages. A further challenge is the performance interference that…

Deep Neural Networks have achieved remarkable achievements across various domains, however balancing performance and generalization still remains a challenge while training these networks. In this paper, we propose a novel framework that…

In this paper, we study a simple and generic framework to tackle the problem of learning model parameters when a fraction of the training samples are corrupted. We first make a simple observation: in a variety of such settings, the…

One concern with the rise of large language models lies with their potential for significant harm, particularly from pretraining on biased, obscene, copyrighted, and private information. Emerging ethical approaches have attempted to filter…

The integration of multi-document pre-training objectives into language models has resulted in remarkable improvements in multi-document downstream tasks. In this work, we propose extending this idea by pre-training a generic multi-document…