mlsanityData to 13 Oct 2026

MaSRead: Content-Addressed Reading of Replicated Latent Stores

arXiv 2608.11218 ↗First posted 21 Jul 2026cs.AI cs.LG cs.MAcited by 0

On the map: Efficient LLM inference and serving → KV cache optimization

Abstract (arXiv)

Independent agents that reason in latent space can share computed state as key-value cache fragments rather than text. Merged by a conflict-free replicated data type, these fragments form a store that converges under any delivery order or duplication. Yet a later query, unknown at encode time, cannot reliably read the merged cache: colocated fragments interfere, so colocation is not addressability. MaSRead addresses the read to content. It routes through opaque keyed tag sets derived from fragment words and decodes each selected fragment under a hard attention mask that hides the rest. Under lexical connectivity, a graph walk reaches the fragments required by a multi-hop query. Across chain, pipeline, symmetric, hub, and natural-language stores, MaSRead recovers visited fragments in isolation, remains effective as unrelated fragments accumulate, and transfers to another model family. After routing, materialized decoding depends on fragment length rather than total store size; end-to-end work still includes store-dependent routing and one read per visited fragment. The limits are explicit: lexical routing can miss disconnected evidence, and answer composition remains bounded by the frozen reader. Thus a replicated latent store becomes selectively readable for later queries when the needed fragments connect to the query through content.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Placed in KV cache optimization, one of 12 directions in Efficient LLM inference and serving. Builds on 33 of its 42 references that are in the map. First posted on arXiv 21 Jul 2026.

Builds on

33 of 42 references.

Closest ancestors · 8

Same direction, newest first

  1. 2026ProphetKV: User-Query-Driven Selective Recomputation for Efficient KV Cache Reuse in Retrieval-Augmented Generation31 Jan
  2. 2025Cache-to-Cache: Direct Semantic Communication Between Large Language Models3 Oct
  3. 2025KVLink: Accelerating Large Language Models via Efficient KV Cache Reuse21 Feb
  4. 2024Model Tells You Where to Merge: Adaptive KV Cache Merging for LLMs on Long-Context Tasks11 Jul
  5. 2024CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusion26 May
  6. 2024SnapKV: LLM Knows What You are Looking for Before Generation22 Apr
  7. 2023Efficient Memory Management for Large Language Model Serving with PagedAttention12 Sep
  8. 2023H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models24 Jun

Other references · 25

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,221
  2. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,716
  3. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,412
  4. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,137
  5. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 902
  6. 2020Hopfield Networks is All You Needcited by 467
  7. 2023Prompt Cache: Modular Attention Reuse for Low-Latency Inferencecited by 258
  8. 2022Memorizing Transformerscited by 222
  9. 2023Landmark Attention: Random-Access Infinite Context Length for Transformerscited by 209
  10. 2025MemGen: Weaving Generative Latent Memory for Self-Evolving Agentscited by 103
  11. 2022Parallel Context Windows for Large Language Modelscited by 81
  12. 2025Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thoughtcited by 73
13 more
  1. 2025Latent Collaboration in Multi-Agent Systemscited by 68
  2. 2024EPIC: Efficient Position-Independent Caching for Serving Large Language Modelscited by 45
  3. 2026The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlookcited by 41
  4. 2025Enabling Agents to Communicate Entirely in Latent Spacecited by 32
  5. 2025APE: Faster and Longer Context-Augmented Generation via Adaptive Parallel Encodingcited by 29
  6. 2024Deliberation in Latent Space via Differentiable Cache Augmentationcited by 25
  7. 2026LatentMem: Customizing Latent Memory for Multi-Agent Systemscited by 14
  8. 2026Agent Primitives: Reusable Latent Building Blocks for Multi-Agent Systemscited by 13
  9. 2026NextMem: Towards Latent Factual Memory for LLM-based Agentscited by 11
  10. 2025CLaRa: Bridging Retrieval and Generation with Continuous Latent Reasoningcited by 10
  11. 2026FlashMem: Distilling Intrinsic Latent Memory via Computation Reusecited by 7
  12. 2026Conflict-Free Replicated Data Types for Neural Network Model Merging: A Two-Layer Architecture Enabling CRDT-Compliant Model Merging Across 26 Strategiescited by 2
  13. 2026Cache Merging as a Convergent Replicated State for Multi-Agent Latent Reasoningcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026DeepRead: Document Structure-Aware Reasoning to Enhance Agentic Search
  2. 2026Persistent Memory in Multi-Agent LLM Inference: What It Costs, What It Buys, and When You Can Tell
  3. 2026Knowing the Store: What a Memory Backend Must Write Down Before an Agent Can Read It
  4. 2026PARSER: Read in Parallel, Reason in Depth for Long-Context LLM Agents
  5. 2025Using Span Queries to Optimize for Cache and Attention Locality
  6. 2025A New HOPE: Domain-agnostic Automatic Evaluation of Text Chunking
  7. 2026RIT-RAG: Navigating Document Corpora with Retrieval-Induced Trees
  8. 2026BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms