MoQAE learns chunk-level KV cache bit-width selection with a mixture-of-experts router, improving accuracy and efficiency over prior quantization methods.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MoQAE: Mixed-Precision Quantization for Long-Context LLM Inference via Mixture of Quantization-Aware Experts
MoQAE learns chunk-level KV cache bit-width selection with a mixture-of-experts router, improving accuracy and efficiency over prior quantization methods.