DepthKV: Layer-Dependent KV Cache Pruning for Long-Context LLM Inference

source_context

{
  "surface": "paper",
  "mode": "paper",
  "query": "DepthKV: Layer-Dependent KV Cache Pruning for Long-Context LLM Inference",
  "normalized_query": "2604.24647",
  "route": "/paper/depthkv-layer-dependent-kv-cache-pruning-for-long-context-llm-inference",
  "paper_ref": "depthkv-layer-dependent-kv-cache-pruning-for-long-context-llm-inference",
  "topic_slug": null,
  "benchmark_ref": null,
  "dataset_ref": null
}