Work

KV Cache Quantization Benchmark: 4-Bit KIVI Finds 30 of 30 Needles

Year: 2026
Type: preprint
Abstract: Research article: KV Cache Quantization Benchmark: 4-Bit KIVI Finds 30 of 30 Needles
Language: English
Open Access status: green
OpenAlex ID: W7234161484
Cites:
Cited by:
Related to: 10