Metaformer: A Transformer That Tends to Mine Metaphorical-Level Information

Article Computer Science, Artificial Intelligence

Dialogue Relation Extraction with Document-Level Heterogeneous Graph Attention Networks

Hui Chen et al.

Summary: We propose a heterogeneous graph attention network to address the problem of dialogue relation extraction. Compared with several popular sequence-based and graph-based models, our method shows superior performance on the benchmark dataset DialogRE.

COGNITIVE COMPUTATION (2023)

Add to Collection

Article Multidisciplinary Sciences

Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences

Alexander Rives et al.

Summary: The deep contextual language model trained through unsupervised learning on protein sequences contains information about biological properties, has a multiscale structural organization, and can be used to improve predictions for protein mutational effects, secondary structure, and long-range contacts.

PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA (2021)

Add to Collection

Proceedings Paper Acoustics

DEVELOPING REAL-TIME STREAMING TRANSFORMER TRANSDUCER FOR SPEECH RECOGNITION ON LARGE-SCALE DATASET

Xie Chen et al.

Summary: This research investigates the potential of Transformer Transducer (T-T) models for low latency and fast decoding on large-scale datasets by combining Transformer-XL and chunk-wise streaming processes, showing that it outperforms other models in streaming scenarios. Runtime cost and latency can be optimized with a relatively small look-ahead.

2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021) (2021)

Add to Collection

Review Computer Science, Artificial Intelligence