-
ACL Findings 20232023Entities can be expressed in diverse formats, such as texts, images, or column names and cell values in tables. While existing entity linking (EL) models work well on per modality configuration, such as text-only EL, visual grounding, or schema linking, it is more challenging to design a unified model for diverse modality configurations. To bring various modality configurations together, we constructed
-
ACL Findings 20232023We propose CHRT (Control Hidden Representation Transformation) — a controlled language generation framework that steers large language models to generate text pertaining to certain attributes (such as toxicity). CHRT gains attribute control by modifying the hidden representation of the base model through learned transformations. We employ a contrastive-learning framework to learn these transformations that
-
ACL 20232023Despite exciting progress in causal language models, the expressiveness of their representations is largely limited due to poor discrimination ability. To remedy this issue, we present CONTRACLM, a novel contrastive learning framework at both the token-level and the sequence-level. We assess CONTRACLM on a variety of downstream tasks. We show that CONTRACLM enhances the discrimination of representations
-
ACL Findings 20232023Self-rationalizing models that also generate a free-text explanation for their predicted labels are an important tool to build trustworthy AI applications. Since generating explanations for annotated labels is a laborious and costly process, recent models rely on large pretrained language models (PLMs) as their backbone and few-shot learning. In this work we explore a self-training approach leveraging both
-
ACL 20232023Despite the popularity of Shapley Values in explaining neural text classification models, computing them is prohibitive for large pretrained models due to a large number of model evaluations. In practice, Shapley Values are often estimated with a small number of stochastic model evaluations. However, we show that the estimated Shapley Values are sensitive to random seed choices — the top-ranked features
Related content
-
July 07, 2022The breadth and originality of Amazon’s natural-language-processing research are on display at the annual meeting of the North American chapter of the Association for Computational Linguistics.
-
June 29, 2022President’s visit part of a mission to preserve the Icelandic language in the digital age.
-
June 28, 2022Amazon’s TabTransformer model is now available through SageMaker JumpStart and the official release of the Keras open-source library.
-
June 22, 2022Rohit Prasad on the pathway to generalizable intelligence and what excites him most about his re:MARS keynote.
-
June 13, 2022Natural Language Processing with AWS AI Services seeks to demystify NLP for just about anyone.
-
June 10, 2022Papers focus on learning previously unseen intents and personalization, both generally and in the specific case of recipe recommendation.