CompressKV uses Semantic Retrieval Heads to guide KV-cache token selection and layer-wise budget allocation, retaining over 97% performance with 3% cache on LongBench QA tasks.
Attention heads of large language models: A survey
8 Pith papers cite this work, alongside 5 external citations. Polarity classification is still indexing.
8
Pith papers citing it
5
external citations · external index
representative citing papers
LVLMs show vocabulary hijacking by inert tokens that decode to hijacking anchors; HABI locates them, NHAR finds resilient heads, and HAVAE boosts those heads to cut hallucinations.
AP-MAE reconstructs masked attention patterns in LLMs with high accuracy, generalizes across models, predicts generation correctness at 55-70%, and enables 13.6% accuracy gains via targeted interventions.
TRACE-RPS drops LLM attribute inference accuracy from around 50% to below 5% via fine-grained anonymization plus a two-stage rejection optimization.
A new framework using Task Subspace Logit Attribution localizes attention heads specialized for task recognition and task learning in in-context learning, showing they align and rotate hidden states within a task subspace.
Reasoning in large output spaces proceeds via shortlisting then fine-grained reasoning; this characterization enables a mechanistic distillation strategy that outperforms standard distillation.
A training-free method that steers specific attention heads toward caption-query states reduces object hallucination in LVLMs by ~6% on average across five benchmarks.
ART replaces uniform attention in shallow LLM layers with local attention patterns to reduce hallucinations across multiple model architectures.
citing papers explorer
-
CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
CompressKV uses Semantic Retrieval Heads to guide KV-cache token selection and layer-wise budget allocation, retaining over 97% performance with 3% cache on LongBench QA tasks.