Cache Merging as a Convergent Replicated State for Multi-Agent Latent Reasoning
摘要
论文提出 Multi-agent latent reasoning 通过将代理的 KV-caches 合并成单个上下文的做法。现有方法 BagMerge 沿序列轴拼接并用 RoPE 重新编码，结果非交换，输入顺序取决于具体场景、潜层步骤预算和模型规模。作者引入 CanonicalMerge，固定布局至内容：按中间层 K-norm 均值排序 caches，算法验证（N<=5）和真实 Qwen3-1.7B / 4B 上 bit-for-bit 证明结果在任何输入置换下均字节相同。进一步将复制状态与解码时布局分离，状态为内容寻址的潜层片段集合，合并为集合并集，形成 CvRDT（commutative, associative, idempotent, absorbing），CanonicalMerge 为其确定性渲染。因渲染字节等价，所有 N = 2 准确率数字不变重复副本被吸收而非重新拼接。在分区域推理基准中，CanonicalMerge 在每种场景 - 预算 - 顺序组合中匹配最佳 BagMerge 顺序，只以微小且统计上不显著的准确率损失换取无条件结构保证。行为在真实多文档 QA（HotpotQA）中可迁移，而最接近的训练免费输出融合基线 PackLLM 在匹配预算下落后 45 分，表明缓存级合并与输出级融合处于不同阶段。最后对于 k> 2 该方法可传输和共置潜层痕迹但不自身组成它们，这被描述为未来工作动机。
荐读理由
提出 Cache Merging 作为 CvRDT 状态：缓存按 K-norm 排序字节等效、去重吸收、多代理 latent reasoning 直接实现，避开 BagMerge 排序依赖
原文
Computer Science > Multiagent Systems
[Submitted on 1 Jul 2026]
Title:Cache Merging as a Convergent Replicated State for Multi-Agent Latent Reasoning
Authors:Carlos Baquero, Luís Brito
View a PDF of the paper titled Cache Merging as a Convergent Replicated State for Multi-Agent Latent Reasoning, by Carlos Baquero and 1 other authors
Abstract:Multi-agent latent reasoning composes agents' KV-caches into one context for a final agent. Prior work (Agent Primitives) does this by concatenating caches along the sequence axis with RoPE re-encoding, which we call BagMerge. BagMerge is non-commutative, and the best input ordering is unpredictable, shifting with the regime, the latent-step budget, and the model scale. We make this exchange a convergent replicated state. First, CanonicalMerge fixes the layout by content: ordering caches by mean K-norm at a middle layer renders the merged cache byte-identical under any input permutation, verified algorithmically (arity N<=5) and bit-for-bit on real Qwen3-1.7B and 4B state. Second, we separate the replicated state from decode-time layout: the state is a set of content-addressed latent fragments whose merge is set union, a state-based CvRDT (commutative, associative, idempotent, absorbing), and CanonicalMerge is its deterministic render. Because the render is byte-equivalent, every N=2 accuracy number carries over unchanged and re-delivered duplicates are absorbed rather than re-concatenated. On a partitioned-reasoning benchmark, CanonicalMerge matches the best BagMerge ordering in every regime-by-budget-by-ordering cell without knowing which order is best, trading a small, statistically insignificant accuracy margin for an unconditional structural guarantee. The behaviour transfers to real multi-document QA (HotpotQA), while the closest training-free output-fusion baseline (PackLLM) loses by 45 points at matched budget, placing cache-level merging in a regime distinct from output-level fusion. Finally, at k>2 the approach transports and colocates latent traces but does not by itself compose them, which we characterize to motivate future work.
https://doi.org/10.48550/arXiv.2607.01308
arXiv-issued DOI via DataCite
| Subjects: | Multiagent Systems (cs.MA) |
|---|---|
| ACM classes: | I.2.7; I.2.11; C.2.4 |
| Cite as: | arXiv:2607.01308 [cs.MA] |
| (or arXiv:2607.01308v1 [cs.MA] for this version) | |
Submission history
From: Carlos Baquero [view email] [v1] Wed, 1 Jul 2026 17:05:04 UTC (55 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Cache Merging as a Convergent Replicated State for Multi-Agent Latent Reasoning, by Carlos Baquero and 1 other authors
Current browse context:
cs.MA
Change to browse by:
References & Citations
Loading...
BibTeX formatted citation
Data provided by:
Bookmark
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
ScienceCast (What is ScienceCast?)
Demos
Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
这条对你有帮助吗？