Skip to content
conftrace_

self-attention matrix

4 papers

Explore in graph

Also known as

SAM

Co-occurring keywords · papers mentioning both

word mover's distance (1) fused gromov-wasserstein distance (1) hierarchical fusion (1) over-trust penalty (1) over-smoothing phenomenon (1) graph perspective (1) flops (1) cuda time (1) rank-based strategy (1) fused gromov-wasserstein (1)

Papers

SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference ICML 2025 OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation CVPR 2024 Improving word mover’s distance by leveraging self-attention matrix EMNLP 2023 Revisiting Over-smoothing in BERT from the Perspective of Graph ICLR 2022