Keys And Value

Follow this topic Ask about this topic

Momentum

8 papers in the last four weeks, up 167% on the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 28

All topics

EchoPress: Query-Agnostic KV Cache Pruning via Virtual Context Reconstruction
Sep 30, 2026Jiawei Lin, Saibo Geng, Thomas BourgeatKey-Value Cache Compression Keys And Value
EPFL
D $^2$ -VLA: Dual-Memory Dual-Frequency Vision-Language-Action Model For Long Dynamic Manipulation
Sep 28, 2026Zijian Ye, Chengqi Wei, Wei Huang +9Keys And Value
The University of Hong Kong · Southern University of Science and Technology
Draft-KV: Learning Useful Latent Communication Between Language Models
Sep 28, 2026Linquan Wu, Shichang Meng, Tianxiang Jiang +7Shared Control Keys And Value
City University of Hong Kong · University of Science and Technology of China · University of Electronic Science and Technology of China +3
Does Execution Require Target KV Fidelity? A Mixed-Fidelity KV Runtime for LLM Serving
Sep 27, 2026Jiantong Jiang, Yue Yang, Peiyu Yang +1Large Language Model Serving Keys And Value
University of Melbourne · Maincode
GSM: Efficient Language Modeling with Shared Global State
Sep 27, 2026Yunao Zheng, Bin Wen, Xiaojie Wang +12Language Modeling Keys And Value
Beijing University of Posts and Telecommunications. · Kuaishou Technology.
Shared Global KV with Layer-Specific Local History
Sep 23, 2026Xinglang XianKeys And Value Large Language Model Decoding
Written as a Record, Read as an Address: What a Forward Pass Leaves in an Operation's KV Cache
Sep 21, 2026Lingfeng Wu, Behzad ShomaliKeys And Value Key-Value Cache
University of Bonn · Lamarr Institute
Shared-Prefix KV Reuse Across Standard LoRA Adapters: Quality and Serving Tradeoffs
Sep 15, 2026Dushyant RajputKv-Cache Management Key-Value Cache
AltSlate Labs LLP
AVERT: Audio-Verified Adjudication for Spoken Dialogue State Tracking
Sep 1, 2026Chunggi Lee, Hanspeter PfisterDialogue Keys And Value
Harvard University Cambridge, Massachusetts, USA
A Universal Context-Reuse Layer for Cross-Model KV Sharing
Aug 31, 2026Yi Li, Dongming Jiang, Yi Zhao +1Kv-Cache Management Keys And Value
University of Texas at Dallas
WhiteMatter: All-to-All Cross-Layer Connections via KV Source Mixing
Aug 19, 2026Wenbo Zhang, Xiang RenTransformer Architectures Cross-Layer Interactions
University of Southern California
SAKI: Score-Aware Low-Rank Key Indexing with Random-Matrix Noise Correction for KV Retrieval
Aug 4, 2026Lin ZhangKey-Value Cache Keys And Value
Data Science and Platform, Cotality
Beyond KV Reconstruction: Functional Reconstruction for MLA Draft Models in Speculative Decoding
Jul 29, 2026Weiye Shi, Fanxu Meng, Muhan ZhangSpeculative Decoding LLM Inference Optimization
Institute for Artificial Intelligence, Peking University
Memory Layer: Train the In-Model Cache for Recommendation Models
Jul 27, 2026Liangyuan Na, Gufan Yin, Yixin Bao +19Text Embeddings Cache
Meta
From Pixels to Pairs: A Comprehensive Benchmark of LLM-Driven Key-Value Extraction in Noisy Document Settings
Jul 14, 2026Zahra AnvariExtraction Keys And Value
Independent AI Researcher.
RaBitQCache: Rotated Binary Quantization for KVCache in Long Context LLM Inference
Jun 30, 2026Wenhao Li, Jinhao Dong, Hailin Zhang +3LLM Inference Optimization Dynamic Sparse Attention
School of Information, Renmin University of China, Beijing, China · Peking University, Beijing, China
Execution-State Capsules: Graph-Bound Execution-State Checkpoint and Restore for Low-Latency, Small-Batch, On-Device Physical-AI Serving
Jun 18, 2026Liang SuKv-Cache Management Intermediate Checkpoints
ReasonAlloc: Hierarchical Decoding-Time KV Cache Budget Allocation for Reasoning Models
Jun 9, 2026Wenhao Liu, Hao Shi, Yunhe Li +7Large Reasoning Models Reasoning Skills
1Tsinghua University · 2City University of Hong Kong · 3Peking University +2
Q-Delta: Beyond Key-Value Associative State Evolution
Jun 7, 2026Sumin Park, Seojin Kim, Noseong ParkKimi Delta Attention Recurrent State
Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Republic of Korea.
P-Cast Precision in FP8 Attention: Sink-Induced Collapse and the Optimality of S=2^8
Jun 2, 2026Reed LauFull-Precision Performance Fp8
Tencent
TGV-KV: Text-Grounded KV Eviction for Vision-Language Models
Jun 2, 2026Jizhihui Liu, Ruizi Han, Miao Zhang +4Recent Vision-Language Models Key-Value Cache Eviction
Harbin Institute of Technology, Shenzhen · Peng Cheng Laboratory
Do Transformers Need Three Projections? Systematic Study of QKV Variants
Jun 1, 2026Ali Kayyam, Anusha Madan Gopal, M Anthony LewisKey-Value Cache Quantization Transformer Architectures
BrainChip Inc., Laguna Hills, CA, USA.
Segment-driven Structural Induction and Semantic Alignment for Heterogeneous Tabular Representation
Jun 1, 2026Woojun Jung, Susik YoonHierarchical Semantic Alignment Induction
Department of Computer Science & Engineering, Korea University, Seoul, South Korea.
The Time is Here for Just-in-Time Systems: Challenges and Opportunities
May 22, 2026Shu Liu, Alexander Krentsel, Shubham Agarwal +8Invariant Synthesis Keys And Value
UC Berkeley · Bespoke Labs
SPHERICAL KV: Angle-Domain Attention and Rate-Distortion Retention for Efficient Long-Context Inference
May 13, 2026Anay Chauhan, Gurucharan Marthi Krishna Kumar, Arion Das +4Key-Value Cache Keys And Value
1Synopsys, India · 2McGill University, Canada · 3IIIT Ranchi, India +4
Sharp Capacity Thresholds in Linear Associative Memory: From Winner-Take-All to Listwise Retrieval
May 6, 2026Nicholas Barnfield, Juno Kim, Eshaan Nichani +2Dense Associative Memory Full-Rank Correlation Matrices
Harvard University · University of California, Berkeley · Princeton University
River-LLM: Large Language Model Seamless Exit Based on KV Share
Apr 20, 2026Yingtao Shen, An ZouInference Latency Large Language Models(Llms
Shanghai Jiao Tong University Shanghai, China
Invertible Query-Key Coupling Composes with Attention Mechanisms
Apr 2, 2026Barak Gahtan, Alex M. BronsteinLinear Attention Keys And Value
Technion – Israel Institute of Technology · ISTA – Institute of Science and Technology Austria

Sep 30, 2026, cs.LGJ/K move · Enter open · S save

EchoPress: Query-Agnostic KV Cache Pruning via Virtual Context Reconstruction

Jiawei Lin, Saibo Geng, Thomas Bourgeat

EPFL

Key-Value Cache Compression Keys And Value

Open paper AskPDF arXiv

KV cache pruning reduces long-context inference memory usage by evicting less important key-value pairs. KVzip estimates importance through context reconstruction: prompting a model to repeat the context chunk by chunk. This achieves strong compression quality at the cost of additional forward passes. Learned approximations reduce this cost but require model-specific training. We analyze how KVzip identifies important cached information and show how to approximate its reconstruction scores using information already computed during prefill. These findings motivate EchoPress, a training-free method that approximates reconstruction attention using queries and keys from standard prefill. For each request, it reconstructs only the first chunk to calibrate importance scores for the remaining context. Experiments on LongBench and RULER with Qwen3-8B and Llama-3.1-8B-Instruct show that EchoPress matches KVzip in task accuracy across eviction ratios from 50% to 90%, while reducing compression overhead by a factor of 1.7-19.6 and total prefill time by a factor of up to 2.9. Code is available at https://github.com/ljwljwljwljw/kvpress/tree/echo-press.

Momentum

Related topics

Latest papers 28