Hamming Attention Distillation binarizes transformer keys and queries to +1/-1 and prunes attention to the top N links, reporting single-point accuracy losses and large simulated hardware savings.
Prime: A novel processing-in-memory architecture for neural net- work computation in reram-based main memory
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Hamming Attention Distillation: Binarizing Keys and Queries for Efficient Long-Context Transformers
Hamming Attention Distillation binarizes transformer keys and queries to +1/-1 and prunes attention to the top N links, reporting single-point accuracy losses and large simulated hardware savings.