Feb 1, 2023 · This paper proposes diffuseq – a diffusion model for sequence-to-sequence text generation tasks. The x and y pairs are concatenated together and sent to the forward …

This paper offers a novel perspective for sequence modeling by formulating it as an IL problem. It tackles the error compounding issue, a significant drawback of sequence modeling, by …

Feb 1, 2023 · Abstract: Sequence generation applications require satisfying semantic constraints, such as ensuring that programs are correct, using certain keywords, or avoiding undesirable …

Jan 28, 2022 · Abstract: A central goal of sequence modeling is designing a single principled model that can address sequence data across a range of modalities and tasks, particularly on …

Jan 22, 2025 · Protein language (or sequence) models, like the popular ESM2, are now widely used tools for extracting evolution-based protein representations and have achieved significant …

Feb 15, 2018 · For the abstractive model, we introduce a decoder-only architecture that can scalably attend to very long sequences, much longer than typical encoder- decoder …

sequence dynamic to a multiple-input multiple-output (MIMO) one. Here, we found 071 that MIMO is particularly suitable for inference, as the extra expressivity allows for more compute during …

ob-lem, we formulate sequence generation as an imitation learning (IL) problem. This allows us to minimize a variety of divergences between the distribution of sequences generated by an …

This paper looks at the problem of sequence modeling, predicting how a sequence will evolve over time. This is a key problem in domains spanning audio, language modeling, music …

Sequence generation applications require satisfying semantic constraints, such as ensuring that programs are correct, using certain keywords, or avoiding undesir-able content.