KV cache quantization
Description: Techniques that represent transformer key-value cache entries with reduced numerical precision to save memory during language-model inference.
Primary topic: Advanced Neural Network Applications (57%)
OpenAlex ID: keywords/kv-cache-quantization
