This paper formalizes KV management as a causal system of three primitives: KV Admission, Selection, and Eviction, and instantiate KV Admission via Write-Gated KV (WG-KV), a lightweight mechanism that learns to predict token utility before cache entry.
Abstract
Long-context LLM inference is bottlenecked by the quadratic attention complexity and linear Key-Value (KV) cache growth. Prior approaches mitigate this via post-hoc selection or eviction but overlook the root inefficiency: indiscriminate token admission. In this paper, we formalize KV management as a causal system of three primitives: KV Admission, Selection, and Eviction. We instantiate KV Admission via Write-Gated KV (WG-KV), a lightweight mechanism that learns to predict token utility before cache entry. By filtering out redundant states early to maintain a compact global cache alongside a sliding local cache, WG-KV significantly reduces memory usage and accelerates both prefill and decode phases. Our results demonstrate that learning what to write is a principled and practical recipe for efficient long-context inference. Code is available at https://github.com/EMCLab-Sinica/WG-KV.
This work proposes DistillCache, a reinforcement learning framework that formulates KV-cache eviction as a sequential decision problem and demonstrates the effectiveness of learned, distribution-aware policies for memory-efficient long-context LLM inference.
The memory usage and decoding latency of LLM inference grow rapidly with context length. To reduce these costs, key-value (KV) cache compression methods selectively retain cached states based on token importance or differences in attention patterns across heads. However, we discover that retrieval capability varies sub...
Xian-Peng Shang, Can-Bin Huang, Jiang Li et al.· 0 citations
Large language model (LLM) inference is often constrained by both computation and memory, especially in offloading-based deployments where model weights are transferred across memory hierarchies during autoregressive decoding. In this setting, reducing the number of executed layers can lower per-token latency while als...
CompKV is introduced, the first compensation-aware sparse attention framework that divides tokens into blocks and explicitly optimizes selection for the downstream compensation mechanism, and shows that the residual left by block-level mean compensation is governed by both block attention mass and within-block logit va...
Zheng-Hong Huang, Rui-Zhe Yao, Dan-Yi Liu et al.· 1 citation· ⚡1
This study investigates the crucial deployment question of whether Cartridges and other CKVs can handle off-context queries, and proposes Cartridges++, simple modifications to cartridges that retain off-context abilities at small or negligible cost.
Sonia Laguna, João Monteiro, Marco Cuturi et al.· 1 citation
Long-context inference with Large Language Models (LLMs) is bottlenecked by the linearly growing memory of the key-value (KV) cache. Existing compression methods reduce the cache through token eviction or approximation, but degrade sharply at aggressive compression budgets. We propose PatchKV, a training-free framework...
Chanryeol Lee, Chanhyuk Lee, Yeonwoo Choi et al.· 1 citation
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduOct 13, 2026
Machine-learning networks can learn without detailed instructions from a designer. Those that do may offer insight into how the brain reorganizes itself during learning.
MIT News · Artificial Intelligence· news.mit.eduSep 29, 2026
Professor Sherry Turkle’s new book, “Artificial Intimacy,” offers a withering critique of chatbots and the antisocial dynamics she believes they encourage.
