conftrace_

attention augmentation

3 papers

Explore in graph

Co-occurring keywords

continual knowledge learning (2) token importance (16) topic segmentation (32) bert-based model (93) attention pattern (45) large language model (13587) neural network (6616) interpretable attention pattern (1) cross-domain adaptation (112) multi-head self-attention (37)

Papers

Train-Attention: Meta-Learning Where to Focus in Continual Knowledge Learning NIPS 2024
Human Guided Exploitation of Interpretable Attention Patterns in Summarization and Topic Segmentation EMNLP 2022
Translation as Cross-Domain Knowledge: Attention Augmentation for Unsupervised Cross-Domain Segmenting and Labeling Tasks EMNLP 2021