conftrace_

self-attention optimization

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

dynamic hidden state (1) vision transformer efficiency (1) self-attention efficiency (1) one-step-ahead attention (1) token fusion (1) decoder-only language model (1) token reduction (1) context compression (1) autoregressive language model (1) model efficiency (1)

Papers

Dodo: Dynamic Contextual Compression for Decoder-only LMs ACL 2024 Multi-criteria Token Fusion with One-step-ahead Attention for Efficient Vision Transformers CVPR 2024