vToken: Token-Level Virtualization for Reclaimable KV Caches
Score
77提出 token 级虚拟化层，弥补块管理与 token 级逐出之间的空白，属于有明确增量的系统设计。
实现完整、实验覆盖多策略与多指标，但缺少与更复杂内存管理方案的横向对比。
兼容现有内核并显著降低策略集成成本，便于在 vLLM 生产环境中快速落地。
KV 缓存优化是当前大模型部署热点，论文方案紧贴社区关注的高并发与内存效率问题。
来自 arXiv AI/CL/LG 论文，作者为 vLLM 生态贡献者，但尚未看出处同行评审与复现验证记录。
Base 85 × T2 Source weight ×0.9 = 77
Curation line 70Selected
Why selected
The paper presents a system-level cache management innovation, covers multiple eviction policies with clear gains, and fills the gap between block-level management and token-granularity eviction.
AI commentary
「What stands out is using a virtualization approach to resolve the granularity mismatch, rather than stacking yet another cache algorithm. For engineers working on inference system internals, it offers a low-intrusion, reusable abstraction that could influence future KV manager designs.」
Quick overview
Large language model serving faces a critical memory bottleneck: the KV cache grows with sequence length and batch size. PagedAttention uses fixed-size memory blocks to reduce allocator-level fragmentation, but recent KV eviction algorithms operate at a token granularity finer than block-level management. This mismatch causes intra-block fragmentation, leaving a large fraction of allocated KV memory unreclaimable. vToken is a lightweight token-level virtualization layer that decouples logical token liveness from physical block placement. It maintains a stable logical token view through token-table indirection and realizes physical reclamation by repacking live tokens asynchronously. The design preserves PagedAttention kernels and CUDA Graph compatibility. Implemented in vLLM and evaluated with H2O, Random, and Scissorhands, vToken reduces retained KV blocks per request by 27.2%--72.3% and improves SLA-constrained throughput by up to 1.37x. Under a constrained active-KV budget, it extends maximum feasible concurrency by up to 2x, while reducing per-policy integration footprint from 500+ lines to under 50.
- vToken decouples logical token liveness from physical block placement via token-table indirection, solving unreclaimable memory caused by intra-block fragmentation.
- Asynchronous repacking of live tokens enables physical reclamation while preserving PagedAttention kernels and CUDA Graph compatibility.
- Integrated with H2O, Random, and Scissorhands in vLLM, vToken reduces retained KV blocks per request by 27.2%--72.3% and improves SLA-constrained throughput by up to 1.37x.
- Per-policy integration footprint drops from 500+ lines to under 50, and constrained active-KV budgets extend maximum feasible concurrency by up to 2x.
Full text
This is an AI-generated Chinese digest; use "View original" for the source text.
Related items
85
FeaturedSetoka: A Benchmark for Hierarchical User Understanding in Personalized Agents over Heterogeneous Data
Setoka：面向异构数据个性化Agent的层级用户理解基准
This paper is well worth attention, especially for researchers and engineers working on personalized agents, dialogue systems, and memory-augmented AI. It not only provides a benchmark but defines levels of user understanding from psychology, revealing systematic failures of current systems at higher levels. If you are designing memory-based agents, this paper reminds you that memory systems need to support cross-temporal, cross-source abstract reasoning, not just retrieval.
85
FeaturedBAND: Bayesian Network Distribution Regression for High-Dimensional Distribution Estimation – Breaking the Curse
BAND：高维分布估计的贝叶斯网络回归方法——打破维度诅咒
This paper is a must-read for anyone working on high-dimensional statistics, probabilistic modeling, and generative AI. It provides a simple yet powerful theoretical answer: sparsity can break the curse of dimensionality, and Bayesian networks offer a natural framework. Compared to black-box methods like VAEs and GANs, BAND is interpretable and provably convergent, especially valuable for industry needs like reliable confidence intervals. If you research time series forecasting or risk management, this paper may change how you approach high-dimensional density estimation.
84
FeaturedCycle-Consistent and Uncertainty-Aware Neural Surrogates for Tokamak Edge Plasmas
循环一致与不确定性感知的托卡马克边缘等离子体神经代理模型
This paper is worth attention because it demonstrates how advanced AI techniques (cycle consistency, uncertainty quantification) directly address a core computational bottleneck in nuclear fusion. For fusion researchers and engineers, it offers a practical tool to speed up simulations by orders of magnitude without sacrificing accuracy; for AI practitioners, it provides an excellent case of self-supervised learning and inverse problem solving in scientific applications.
83
FeaturedEquivariant Learning of a Transferable Three-Dimensional Classical Density Functional
等变学习可迁移的三维经典密度泛函
The most striking aspect of this paper is its direct coupling of machine learning to a core problem in statistical mechanics, using equivariant networks to learn a transferable density functional without any labels. For researchers in liquid physics or soft matter, this could be a new paradigm: do we need to run separate simulations for every new system anymore? For AI researchers, it also shows how to embed physical symmetries and variational principles into models rather than just fitting data. I recommend focusing on the methods, especially how equivariance and variational consistency are implemented—that's the real essence.
83
FeaturedComputational and Statistical Guarantees of the c-Rectified Flow
c-Rectified Flow 的计算与统计保证
This paper moves rectified flow from 'works in practice' to 'theoretically controllable,' which is especially important for those working on the underlying algorithms of generative models. If you are tuning pipelines like FLUX or SD3, understanding that ordinary rectified flow can fail when covariances do not commute may help you avoid hidden training pitfalls.
83
FeaturedLearning Deployable Manipulation Policies from High-Fidelity UMI Data Alone
从高保真UMI数据学习可部署操作策略
The most striking point of this paper is proving that pure robot-free data can match real teleoperation performance, breaking the convention of needing a small real-robot anchor for post-training. For research groups looking to reduce hardware costs and industrial robot application developers, HiFi-UMI's data production system and open-source dataset are directly usable resources. Pay attention to the engineering details of head-mounted SLAM and microsecond synchronization—they are key to achieving the reported accuracy.
83
FeaturedCapital Markets LLM Reliability Score: From Plausible to Bankable
资本市场LLM可靠性评分：从“看似合理”到“可银行化”
This paper is valuable for both finance and AI communities: it provides an actionable, workflow-oriented evaluation framework that's closer to real needs than generic benchmarks. If you're deploying LLMs in fintech or responsible for explainability/auditability, this is a must-read. The finding on decision usefulness is particularly striking—judges agree it's critical, yet models vary the most on it.
83
FeaturedWhere Should Optimizer State Live? Tiered State Allocation for Memory-Efficient Mixture-of-Experts Training
优化器状态应放在哪里？面向内存高效MoE训练的分层状态分配
This paper is highly valuable for teams training MoE models: it demonstrates that with careful optimizer state allocation, one can dramatically reduce memory without sacrificing accuracy—and even improve performance. It is especially relevant for labs or enterprises constrained by 40 GB or 80 GB GPU memory. The tiered strategy and ablation on momentum retention are worth studying for optimizing training frameworks like DeepSpeed or Megatron.