EchoKV: Efficient KV Cache Compression via Similarity-Based Reconstruction
cs.CL
Submitted: 2026-03-24
Updated: 2026-11-03
Code: https://github.com/noforit/EchoKV
Terminology
Sources
- SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models
- Keep the Cost Down: A Review on Methods to Optimize LLM' s KV-Cache Consumption
- DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
- Palu: Compressing KV-Cache with Low-Rank Projection
- CommonKV: Compressing KV Cache with Cross-layer Parameter Sharing
- xKV: Cross-Layer KV-Cache Compression via Aligned Singular Vector Extraction
- ProxyAttn: Guided Sparse Attention via Representative Heads
- Accelerating Prefilling for Long-Context LLMs via Sparse Pattern Sharing
- RULER: What's the Real Context Size of Your Long-Context Language Models?
- QAQ: Quality Adaptive Quantization for LLM KV Cache
- KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache
- Fast Transformer Decoding: One Write-Head is All You Need
- GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints
- TransMLA: Multi-Head Latent Attention Is All You Need
- X-EcoMLA: Upcycling Pre-Trained Attention into MLA for Efficient and Extreme KV Compression
- Lookahead Q-Cache: Achieving More Consistent KV Cache Eviction via Pseudo Query
- Judge Q: Trainable Queries for Optimized Information Retention in KV Cache Eviction
- ThinK: Thinner Key Cache by Query-Driven Pruning
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data