inference/research
verifiedhigh · 0.90updated 2026-10-03

The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits

paper

Fields

id
sources/quant-bitnet-paper
URL
https://arxiv.org/abs/2402.17764
Publisher
arXiv
Kind
paper
Published
2024-02-27
Accessed
2026-10-03
notes
Microsoft. BitNet b1.58: every weight ternary {-1,0,1}. Abstract read directly; marked 'Work in progress' by the authors.

Notes

Microsoft. BitNet b1.58: every weight ternary {-1,0,1}. Abstract read directly; marked 'Work in progress' by the authors.

Cross-references

This record references no other records.