Abstract The current paradigm for safety alignment of large language models (LLMs) follows a one-size-fits-all approach: the model refuses to interact with any content deemed unsafe by the model provider. This approach lacks flexibility in[…]

Read More

Abstract Numerous capability and safety techniques of Large Language Models (LLMs), including RLHF, automated red-teaming, prompt engineering, and infilling, can be cast as sampling from an unnormalized target distribution defined by a given reward or[…]

Read More

Abstract Recent years saw the rise of multilingual language models that achieve high levels of performance for a large number of tasks, with some of them handling hundreds of languages at once. Premodern languages are[…]

Read More

Abstract In the age of data- and capital-driven machine learning, the gap between technological advancements for high- and low-resource language varieties keeps growing, leaving many with the greatest need for language technologies without access to[…]

Read More

Abstract SoloAudio is an innovative diffusion-based generative model designed for target sound extraction (TSE). It trains latent diffusion models on audio, replacing the previous U-Net backbone with a skip-connected Transformer that operates on latent features.[…]

Read More

Abstract Just as instructions have enabled complex user tasks and zero-shot optimization via prompts, so can instructions in retrieval enable complex information needs with zero-shot adaptation. In this talk, we describe efforts to build a[…]

Read More

Abstract Language model pretraining has been a remarkably strong recipe for cross-task and cross-domain generalization in NLP. However, these gains have come at the expense of control: we rarely control the training data for language[…]

Read More