KV Cache Compression Through the Lens of Transform Coding
The key-value (KV) cache stores information from past tokens and is a major memory bottleneck in long-context inference. Existing quantization methods address this bottleneck by representing the KV cache uniformly with lower-precision data types and designing quantization schemes to minimize reconstruction error in the cache itself, without accounting for how that error propagates through attention mechanisms. We prove that, under a white-noise quantization model, the expected attention-aware distortion decomposes into additive key and value contributions that factor across tokens and channels
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 53%Breakthrough in long-context efficiency announced →
- FuzzySimilar title/name (fuzzy) · 87%LMCache/LMCache →
“Fuzzy title match (0.94): “KV Cache Compression Through the Lens of Transform Coding” ≈ “LMCache/LMCache””
- FuzzyOverlapping authors or contributors · 62%bytedance/deer-flow →
“Shared author/contributor keys: wang”
- FuzzyOverlapping authors or contributors · 62%ray-project/ray →
“Shared author/contributor keys: wang”
- LinkedLinked via arxiv author · 85%Hannah Laus →
“KV Cache Compression Through the Lens of Transform Coding”
- LinkedLinked via arxiv author · 85%Claudio Mayrink Verdun →
“KV Cache Compression Through the Lens of Transform Coding”
- LinkedLinked via arxiv author · 85%Yihao Wang →
“KV Cache Compression Through the Lens of Transform Coding”
- LinkedLinked via arxiv author · 85%Flavio du Pin Calmon →
“KV Cache Compression Through the Lens of Transform Coding”