A dynamic block floating-point format with pivot-focus and adaptive grouping, plus a hierarchical lookup table, lets attention Softmax run in integer-only hardware with negligible accuracy loss.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Pushing the Limits of BFP on Narrow Precision LLM Inference
A dynamic block floating-point format with pivot-focus and adaptive grouping, plus a hierarchical lookup table, lets attention Softmax run in integer-only hardware with negligible accuracy loss.