Distributed Retrieval-Augmented Generation

Chenhao Xu; Longxiang Gao; Xi Zheng; Yuan Miao

REVIEW 4 major objections 3 minor 38 references

Distributed Retrieval-Augmented Generation

T0 review · 4 major / 3 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A topic-aware random walk lets a distributed RAG system match centralized accuracy at roughly half the communication cost.

desk verdict A legitimate decentralized RAG proposal whose headline message savings are real but only demonstrated under a favorable, unstated knowledge distribution; deserving of peer review with revisions. read the letter →

arxiv 2505.00443 v1 pith:ILTXK5WZ submitted 2025-05-01 cs.DC

Chenhao Xu , Longxiang Gao , Yuan Miao , Xi Zheng This is my paper · ORCID

classification cs.DC

keywords retrieval-augmentedgenerationdistributedsystemspeer-to-peernetworkstopic-awarerandomwalkedgecomputinglargelanguagemodelsknowledgeretrievalprivacy-preservingAI

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that retrieval-augmented generation does not need a central knowledge base. It proposes a peer-to-peer framework, DRAG, in which each edge device keeps its own private knowledge base and LLM, and a Topic-Aware Random Walk (TARW) routes the query toward peers likely to hold the answer. Across MMLU, medical, and news benchmarks, TARW retrieves relevant knowledge for 96 to 98 percent of queries and lands F1 scores within roughly two to four points of centralized RAG, while sending fewer messages than flooding at 20 peers and close to half as many at 100 peers. A reader should care because the same architecture removes the central store, which is a privacy and scaling bottleneck, at a modest and partly tunable accuracy cost.

What carries the argument

The object that carries the argument is the TARW algorithm (Algorithm 1), a bounded random walk whose steps are reweighted by topic. At the querying peer, the local LLM extracts topics from the query; each visited peer computes a local retrieval relevance score; if that score clears the threshold theta = 0.8, the peer's knowledge is privacy-filtered and returned, and the peer is recorded in an expertise cache under the query's topics. If no peer can answer, the current peer ranks unvisited neighbors by the cached topic-expertise match and forwards the query to the top k, with a max hop count of 6. The cache is the mechanism that turns raw exploration into learned routing: as successful queries accumulate, message counts converge downward, and the sensitivity experiments show the accuracy jump comes mainly from moving the neighbor-selection count from k=2 to k=4.

What would settle it

Re-run the MMLU, medical, and news experiments holding the network and parameters fixed while varying the per-peer knowledge split from high topic duplication to strict one-topic-per-peer silos. If TARW's hit rate and message savings weaken substantially in the silo condition, the near-centralized claim is an artifact of knowledge overlap rather than a property of topic-aware routing.

Watch

Extended reading notes

Core claim

The central claim is that a decentralized RAG system can come close to the accuracy of a centralized one if knowledge discovery is topic-guided rather than blind. The paper's experiments show DRAG-TARW reaching a 98.11 percent hit rate and 83.92 F1 on MMLU against 99.81 percent and 85.75 for centralized RAG; on medical data, 98.67 percent hit rate and 90.58 F1 against 99.81 percent and 94.46; on news, 96.86 percent and 74.73 against 99.16 percent and 76.66. Meanwhile the blind random-walk baseline retrieves the right knowledge only 21 to 28 percent of the time, and its F1 collapses as the network grows. The conclusion the paper draws is that the distributed setting itself does not cause the accuracy loss; the routing policy does, and a topic-aware policy with a learned peer-expertise cache keeps the loss small while cutting communication.

Load-bearing premise

The load-bearing premise is that relevant knowledge for almost every query sits within a few hops of the peer asking; the paper never says how it splits the datasets into per-peer knowledge bases, so a favorable split that duplicates topics across many peers could make TARW look better than flooding than a disjoint split would.

Editorial extensions

If this is right

DRAG-TARW's F1 stays within 1.8 to 3.9 points of centralized RAG on the three benchmarks, while flooding achieves similar accuracy at notably higher message counts.
At 100 peers the communication saving approaches half: 27.67 messages versus 53.34 on MMLU and 23.71 versus 48.83 on News.
Retrieval without topic guidance fails in this setting: the random-walk baseline's hit rate is 21.96 to 27.62 percent and falls to 3.57 F1 on MMLU at 100 peers.
Against incomplete central knowledge bases, DRAG is more resilient: when the central base is cut to 70 percent of its samples, centralized F1 on MMLU drops to 61.54 percent while DRAG-TARW holds 83.92 percent.
The system gets cheaper over time: on MMLU with 100 peers, average messages per query fall from 34.73 at 500 queries to 27.67 at 10,000 as the peer-expertise cache fills.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The privacy benefit is architectural: no central server holds all data, but shared snippets are not encrypted or anonymized by construction, so a realistic deployment would need to pair DRAG with differential privacy or traffic-obfuscation mechanisms.
Because the advantage rests on topic overlap among peers, the natural stress test is a silo split where each topic lives on exactly one peer; the paper's own k=2 result (F1 56.27 on MMLU) suggests the near-centralized performance may not transfer to disjoint expert networks.
The paper counts each visited peer as one message, so real deployments also pay for routing-table maintenance, cache freshness, and retries; the half-message saving is a lower bound on communication overhead.
Treating the LLM as the router is a transferable idea: topic extraction plus an expertise cache could be applied to hierarchical federated search or multi-hop retrieval over private document collections where the bottleneck is finding who holds what without a central index.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear colleague,

The short version: this is a solid systems paper, not a breakthrough. It puts together decentralized RAG with a topic-aware random walk and an expertise cache in a way I haven't seen in the federated RAG or query-routing literature. The evaluation is broad enough to support the main qualitative claim: on three datasets, TARW lands within a few F1 points of centralized RAG and sends fewer messages than flooding. That is a real, useful result for edge applications that don't want a central knowledge base.

What it does well: the algorithm is clearly described, the baselines are reasonable (CRAG, NoRAG, random walk, flooding), and the sensitivity analysis covers network size, connectivity, model choice, and the k parameter. Figure 4 shows a nice convergence effect from caching. The authors also discuss privacy and reliability limitations honestly in Section 3.3.

The soft spots are about evidence quality, not the core idea. The biggest one, which the stress-test flags correctly, is that the knowledge distribution across peers is never specified. We don't know how the datasets are partitioned, how queries and answers are assigned to peers, or how many peers hold a given snippet. With hit rates around 98% and average hops under 2 on a 20-peer network, the split must be quite favorable—relevant knowledge is probably replicated or close to the querying peer. If the split were adversarial or sparse, TARW's message savings would shrink and perhaps disappear. That assumption needs to be stated and varied experimentally. Second, there are no variance bars; all numbers look like single-run values. Third, the abstract's "half as many messages" claim is true only at 100 peers (about 50% on MMLU and News), not at the default 20 peers where the savings are 37%, 9%, and 29%. That's not fatal, but the headline overstates the default configuration. Fourth, the comparison in Figure 2 against intentionally incomplete CRAG variants is a bit stacked—of course a centralized system with half its knowledge does worse. It still makes a fair point about resilience, but it's not the strongest evidence for DRAG.

None of this breaks the paper. The central argument—that topic-guided routing can approximate centralized retrieval at lower communication cost in a distributed setting—holds up under the stated favorable split. The next revision should disclose the partitioning scheme, test at least one alternative split, and run multiple seeds.

My recommendation: send it to peer review. It deserves a serious referee, and with a revision that closes the knowledge-distribution gap, it would be a useful contribution to the distributed LLM literature.

Referee Report

4 major / 3 minor

Summary. The paper proposes Distributed Retrieval-Augmented Generation (DRAG), a framework that removes the centralized knowledge base from RAG by storing knowledge across peer devices and routing queries with a Topic-Aware Random Walk (TARW). TARW extracts query topics with a local LLM, uses a relevance threshold to select knowledgeable peers, and maintains a cache of peer expertise. The authors compare DRAG-TARW against centralized RAG, random walk, and flooding baselines on MMLU, Medical Extended, and News datasets, reporting near-centralized accuracy with reduced message counts, scaling behavior of message overhead, and sensitivity to network size, connectivity, LLM choice, and the neighbor-selection parameter.

Significance. If the empirical claims hold, DRAG would be a useful step toward privacy-preserving and scalable RAG on edge networks, and the TARW routing idea is simple and plausibly transferable. The manuscript's strengths include a clearly formulated algorithm, open-source code, evaluation across three datasets, three open-weight LLMs, and multiple ablation experiments. However, the core quantitative claim depends on an underspecified knowledge-distribution setup, and the reported results lack statistical variation, so the current evidence is conditional rather than conclusive.

major comments (4)

[§4.1 (Datasets), Algorithm 1 lines 10–13] The paper never specifies how each dataset is partitioned into per-peer knowledge bases, how queries and ground-truth pairs are assigned to peers, how many peers hold a given answer snippet, or which embedding/relevance model implements Relevance(). The reported hit rates (e.g., 98.11% on MMLU) and message counts assume that relevant knowledge is reachable within H_max=6 hops; if relevant snippets are replicated across many peers or if querying peers are close to answer-holding peers, TARW's advantage over flooding is inflated. Because this distribution model is the load-bearing assumption behind Table 1 and Figure 3, the authors must specify the partitioning and replication strategy, then vary it (e.g., different duplication factors, adversarial query-to-peer assignments) to show the result is robust.
[Abstract and §4.3 (Figure 3)] The abstract states that DRAG-TARW uses 'half as many messages as flooding.' At the default 20-peer setting in Table 1, the reductions over flooding are roughly 37% on MMLU (6.87 vs 10.91), 9% on Medical (8.82 vs 9.72), and 29% on News (7.82 vs 10.99); the approximately 50% reduction appears only at 100 peers in Figure 3. The abstract therefore overstates the headline result. The paper should either soften the claim to specify the scaling regime or present a rigorous scaling analysis showing when the 50% reduction holds.
[Table 1, Figures 3–7] All results are single-run point estimates without variance, confidence intervals, or significance tests. Given stochasticity in Barabási-Albert graph generation, query assignment, and LLM generation, the observed differences (e.g., EM 83.90% for DRAG-TARW vs 85.73% for CRAG on MMLU; message counts differing by less than a message on Medical) cannot be assessed for reliability. At minimum, the authors should run multiple seeds and report mean ± std over independently generated networks and knowledge assignments.
[Figure 2] The comparison against centralized RAG variants with incomplete knowledge (CRAG-0.7S, CRAG-0.5S, CRAG-0.7T, CRAG-0.5T) is not controlled. In those variants, the centralized knowledge base is deliberately reduced to 50–70% of snippets or topics, whereas DRAG has access to the full distributed knowledge across all peers. DRAG's superior F1 score in that comparison is therefore expected by construction and does not by itself demonstrate resilience to incomplete knowledge. To support the resilience claim, the authors must apply a comparable completeness constraint to the distributed setting (e.g., remove a fraction of topics or snippets from all peers) and show DRAG still outperforms the equally limited centralized baseline.

minor comments (3)

[§4.1, Algorithm 1] The relevance function used in Line 10–11 is never defined beyond 'semantic similarity'; please state which embedding model (e.g., sentence-transformers or the local LLM itself) computes Relevance(K_i, q), and report the prompt/temperature used for topic extraction.
[§4.1, Medical Extended dataset] The Medical Extended dataset is described as synthetic; please clarify what constitutes a peer's knowledge base for this dataset and how patient-related snippets are split among peers, since this affects the privacy interpretation.
[Throughout] The paper uses placeholder publication venue and DOI fields in the ACM reference format; these should be completed or removed if the manuscript is intended for arXiv distribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: TARW's performance and message savings are empirical benchmark results, not reductions to fitted inputs or self-citations.

full rationale

The paper's central claims are empirical: DRAG-TARW is compared against CRAG, DRAG-FL, DRAG-RW, and NoRAG on public benchmarks. The algorithm parameters (H_max=6, k=4, theta=0.8) are fixed hyperparameters, not fitted to the reported metrics; sensitivity analyses in Figures 3, 5, and 7 explore them after the fact. The routing mechanism (topic extraction, expertise cache, top-k neighbor selection) is a concrete algorithm whose hit rate and message counts depend on the simulated network and knowledge distribution, so the outcome is not forced by definition. The only self-reference is citation [37] in the incentive-mechanisms discussion, which is not load-bearing for the performance or communication results. The unstated dataset-partitioning scheme noted by reviewers is a reproducibility/validity concern about external validity, not a circularity: nothing in the paper defines the target metric in terms of the inputs. Therefore no step reduces a reported result to its own assumptions.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The reported numbers depend on hand-set algorithm parameters (H_max=6, k=4, theta=0.8) and on an unstated dataset-to-peer knowledge distribution. None of these are fitted to data in a formal sense, but they are chosen without full justification and can move the headline F1 and message counts. The central statistical evidence is a single-run simulation with no error bars. No unexplained physical entities are introduced.

free parameters (3)

H_max (maximum hop count) = 6
Set by hand in Section 4.1; bounds how far TARW searches and directly affects hit rate and message count.
k (neighbor selection count) = 4 (default)
Set to 4 in Section 4.1; Figure 7 shows k=2 drops F1 sharply (56.27% vs 83.92% at 20 peers), so the default choice materially shapes results.
theta (relevance threshold) = 0.8
Set in Section 4.1 and used in Algorithm 1 lines 10-13; determines whether a peer is considered to have answered, and therefore hit rate and routing behavior.

assumptions (4)

domain assumption The semantic relevance function with threshold theta=0.8 reliably signals whether a peer can answer the query.
Algorithm 1 lines 10-13 and Section 4.1; the routing decisions and hit-rate metric depend on this signal.
ad hoc to paper The dataset can be split into per-peer knowledge bases such that relevant knowledge for each query is reachable within H_max hops.
Section 4.1 never specifies the split; if this fails, hit rate and message counts in Table 1 would change.
domain assumption Barabási-Albert random graphs adequately model real edge and P2P networks for evaluating DRAG.
Section 4.1 Network Configuration; real networks have different degree distributions, clustering, churn, and failures.
domain assumption LLM topic extraction correlates with the topics of peers' future contributions, and the expertise cache improves routing.
Section 3.2, Algorithm 1 lines 1 and 16; the mechanism's benefit is measured only through this correlation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Distributed Retrieval-Augmented Generation." pith.science (2026). https://pith.science/paper/ILTXK5WZ

@misc{pith2026250500443,
  author       = {Pith},
  title        = {Pith review of: Distributed Retrieval-Augmented Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ILTXK5WZ}},
  note         = {Machine review of arXiv:2505.00443}
}

read the original abstract

As large language models (LLMs) become increasingly adopted on edge devices, Retrieval-Augmented Generation (RAG) is gaining prominence as a solution to address factual deficiencies and hallucinations by integrating external knowledge. However, centralized RAG architectures face significant challenges in data privacy and scalability. For instance, smart healthcare services often rely on collecting sensitive patient data and building a centralized knowledge base to provide better diagnosis and treatment advice, while privacy concerns significantly impede this process. Besides, maintaining a comprehensive and continuously updated knowledge base is costly, particularly in response to regional epidemics and rapidly mutating viruses. To address these challenges, this paper introduces Distributed Retrieval-Augmented Generation (DRAG), a novel framework that improves data privacy by eliminating the need for a centralized knowledge base and restoring data control to owners. DRAG incorporates a Topic-Aware Random Walk (TARW) algorithm that leverages LLMs to extract query topics and facilitate targeted peer discovery within a peer-to-peer network, enabling efficient knowledge retrieval in decentralized environments. Extensive experiments across three diverse datasets and LLMs demonstrate that DRAG with TARW achieves near-centralized RAG performance by using half as many messages as flooding. The code is available at https://github.com/xuchenhao001/DRAG.

Figures

Figures reproduced from arXiv: 2505.00443 by the authors.

**Figure 2.** Comparative F1 Scores of DRAG and Centralized RAG Variants Under Varying Knowledge Base Completeness. CRAG-0.7S means the centralized knowledge base contains 70% of knowledge snippets, while CRAG-0.7T means it includes 70% of topics. • Precision: The ratio of correctly predicted tokens to the total number of predicted tokens, measuring the quality of the generated output. • Recall: The ratio of correctly predicted … view at source ↗

**Figure 3.** Impact of network size on F1 score (first row) and message overhead (second row) in DRAG. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 7.** Figure 7: Influence of the neighbor selection count, [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

**Figure 5.** Figure 5: Impact of peer connectivity on performance [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

**Figure 6.** Figure 6: Influence of large language model on DRAG [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 19 canonical work pages

[1]

A review on edge large language models: Design, execution, and applications,

Y. Zheng, Y. Chen, B. Qian, X. Shi, Y. Shu, and J. Chen, “A review on edge large language models: Design, execution, and applications, ” ACM Computing Surveys, 2024

work page 2024
[2]

Mobile edge intelligence for large language models: A contemporary survey,

G. Qu, Q. Chen, W. Wei, Z. Lin, X. Chen, and K. Huang, “Mobile edge intelligence for large language models: A contemporary survey, ”IEEE Communications Surveys & Tutorials, 2025

work page 2025
[3]

Pre-trained language models in biomedical domain: A systematic survey,

B. Wang, Q. Xie, J. Pei, Z. Chen, P. Tiwari, Z. Li, and J. Fu, “Pre-trained language models in biomedical domain: A systematic survey, ”ACM Computing Surveys, vol. 56, no. 3, pp. 1–52, 2023

work page 2023
[4]

Vision language models in autonomous driving: A survey and outlook,

X. Zhou, M. Liu, E. Yurtsever, B. L. Zagar, W. Zimmer, H. Cao, and A. C. Knoll, “Vision language models in autonomous driving: A survey and outlook, ”IEEE Transactions on Intelligent Vehicles, 2024

work page 2024
[5]

Survey of hallucination in natural language generation,

Z. Ji, N. Lee, R. Frieske, T. Yu, D. Su, Y. Xu, E. Ishii, Y. J. Bang, A. Madotto, and P. Fung, “Survey of hallucination in natural language generation, ” ACM computing surveys, vol. 55, no. 12, pp. 1–38, 2023

work page 2023
[6]

Retrieval-augmented generation for large language models: A survey,

Y. Gao, Y. Xiong, X. Gao, K. Jia, J. Pan, Y. Bi, Y. Dai, J. Sun, and H. Wang, “Retrieval-augmented generation for large language models: A survey, ” arXiv preprint arXiv:2312.10997, 2023

arXiv 2023
[7]

A survey on rag meeting llms: Towards retrieval-augmented large language models,

W. Fan, Y. Ding, L. Ning, S. Wang, H. Li, D. Yin, T.-S. Chua, and Q. Li, “A survey on rag meeting llms: Towards retrieval-augmented large language models, ” inProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , 2024, pp. 6491–6501

work page 2024
[8]

Don’t forget private retrieval: distributed private similarity search for large language models,

G. Zyskind, T. South, and A. Pentland, “Don’t forget private retrieval: distributed private similarity search for large language models, ” inPro- ceedings of the Fifth Workshop on Privacy in Natural Language Process- ing. Bangkok, Thailand: Association for Computational Linguistics, Aug. 2024, pp. 7–19

work page 2024

Show all 38 references

[9]

Privacy-preserving retrieval augmented generation with differential privacy,

T. Koga, R. Wu, and K. Chaudhuri, “Privacy-preserving retrieval augmented generation with differential privacy, ” arXiv preprint arXiv:2412.04697, 2024

2024
[10]

Cache me if you can: The case for retrieval augmentation in federated learning,

A. Muhamed, P. Thaker, M. T. Diab, and V. Smith, “Cache me if you can: The case for retrieval augmentation in federated learning, ” inPrivacy Regulation and Protection in Machine Learning , 2024. Distributed Retrieval-Augmented Generation Conference acronym ’XX, June 03–05, 20...

2024
[11]

Mitigating the privacy issues in retrieval-augmented gen- eration (rag) via pure synthetic data,

S. Zeng, J. Zhang, P. He, J. Ren, T. Zheng, H. Lu, H. Xu, H. Liu, Y. Xing, and J. Tang, “Mitigating the privacy issues in retrieval-augmented gen- eration (rag) via pure synthetic data, ”arXiv preprint arXiv:2406.14773, 2024

2024 arXiv
[12]

Evaluating the efficacy of open-source llms in enterprise-specific rag systems: A comparative study of performance and scalability,

A. Purwar et al. , “Evaluating the efficacy of open-source llms in enterprise-specific rag systems: A comparative study of performance and scalability, ”arXiv preprint arXiv:2406.11424, 2024

2024 arXiv
[13]

Benchmarking large language models in retrieval-augmented generation,

J. Chen, H. Lin, X. Han, and L. Sun, “Benchmarking large language models in retrieval-augmented generation, ” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 16, 2024, pp. 17 754– 17 762

2024
[14]

Retrieval-augmented generation for knowledge-intensive nlp tasks,

P. Lewis, E. Perez, A. Piktus, F. Petroni, V. Karpukhin, N. Goyal, H. Küt- tler, M. Lewis, W.-t. Yih, T. Rocktäschelet al., “Retrieval-augmented generation for knowledge-intensive nlp tasks, ” Advances in neural information processing systems, vol. 33, pp. 9459–9474, 2020

2020
[15]

Retrieval aug- mented language model pre-training,

K. Guu, K. Lee, Z. Tung, P. Pasupat, and M. Chang, “Retrieval aug- mented language model pre-training, ” inInternational conference on machine learning. PMLR, 2020, pp. 3929–3938

2020
[16]

Dense passage retrieval for open-domain question answering,

V. Karpukhin, B. Oguz, S. Min, L. Wu, S. Edunov, D. Chen, and W.-t. Yih, “Dense passage retrieval for open-domain question answering, ” in Proceedings of the 2020 Conference on Empirical Methods in Natu- ral Language Processing (EMNLP) . Association for Computational Linguisti...

2020
[17]

Self-rag: Learning to retrieve, generate, and critique through self-reflection,

A. Asai, Z. Wu, Y. Wang, A. Sil, and H. Hajishirzi, “Self-rag: Learning to retrieve, generate, and critique through self-reflection, ” inThe Twelfth International Conference on Learning Representations , 2023

2023
[18]

From local to global: A graph rag approach to query-focused summarization,

D. Edge, H. Trinh, N. Cheng, J. Bradley, A. Chao, A. Mody, S. Truitt, and J. Larson, “From local to global: A graph rag approach to query-focused summarization, ”arXiv preprint arXiv:2404.16130, 2024

2024 arXiv
[19]

Grag: Graph retrieval-augmented generation,

Y. Hu, Z. Lei, Z. Zhang, B. Pan, C. Ling, and L. Zhao, “Grag: Graph retrieval-augmented generation, ”arXiv preprint arXiv:2405.16506, 2024

2024 arXiv
[20]

Gnn-rag: Graph neural retrieval for large language model reasoning,

C. Mavromatis and G. Karypis, “Gnn-rag: Graph neural retrieval for large language model reasoning, ”arXiv preprint arXiv:2405.20139, 2024

2024 arXiv
[21]

Rankrag: Unifying context ranking with retrieval- augmented generation in llms,

Y. Yu, W. Ping, Z. Liu, B. Wang, J. You, C. Zhang, M. Shoeybi, and B. Catanzaro, “Rankrag: Unifying context ranking with retrieval- augmented generation in llms, ”Advances in Neural Information Pro- cessing Systems, vol. 37, pp. 121 156–121 184, 2025

2025
[22]

Large language model routing with benchmark datasets,

T. Shnitzer, A. Ou, M. Silva, K. Soule, Y. Sun, J. Solomon, N. Thompson, and M. Yurochkin, “Large language model routing with benchmark datasets, ” inFirst Conference on Language Modeling , 2024

2024
[23]

Un- supervised query routing for retrieval augmented generation,

F. Mu, L. Zhang, Y. Jiang, W. Li, Z. Zhang, P. Xie, and F. Huang, “Un- supervised query routing for retrieval augmented generation, ”arXiv preprint arXiv:2501.07793, 2025

2025 arXiv
[24]

Hi- erarchical retrieval-augmented generation model with rethink for multi-hop question answering,

X. Zhang, M. Wang, X. Yang, D. Wang, S. Feng, and Y. Zhang, “Hi- erarchical retrieval-augmented generation model with rethink for multi-hop question answering, ”arXiv preprint arXiv:2408.11875, 2024

2024 arXiv
[25]

Archrag: Attributed community-based hierarchical retrieval-augmented generation,

S. Wang, Y. Fang, Y. Zhou, X. Liu, and Y. Ma, “Archrag: Attributed community-based hierarchical retrieval-augmented generation, ”arXiv preprint arXiv:2502.09891, 2025

2025 arXiv
[26]

Feb4rag: Evaluat- ing federated search in the context of retrieval augmented generation,

S. Wang, E. Khramtsova, S. Zhuang, and G. Zuccon, “Feb4rag: Evaluat- ing federated search in the context of retrieval augmented generation, ” in Proceedings of the 47th International ACM SIGIR Conference on Re- search and Development in Information Retrieval , 2024, pp. 763–773

2024
[27]

C-fedrag: A confidential federated retrieval-augmented generation system,

P. Addison, M.-T. H. Nguyen, T. Medan, M. T. Manzari, B. McElrone, L. Lalwani, A. More, S. Sharma, H. R. Roth, I. Yang et al., “C-fedrag: A confidential federated retrieval-augmented generation system, ”arXiv preprint arXiv:2412.13163, 2024

2024 arXiv
[28]

The good and the bad: Exploring privacy issues in retrieval-augmented generation (rag),

S. Zeng, J. Zhang, P. He, Y. Xing, Y. Liu, H. Xu, J. Ren, S. Wang, D. Yin, Y. Chang et al., “The good and the bad: Exploring privacy issues in retrieval-augmented generation (rag), ”arXiv preprint arXiv:2402.16893, 2024

2024 arXiv
[29]

Rag-thief: Scalable extraction of private data from retrieval-augmented generation ap- plications with agent-based attacks,

C. Jiang, X. Pan, G. Hong, C. Bao, and M. Yang, “Rag-thief: Scalable extraction of private data from retrieval-augmented generation ap- plications with agent-based attacks, ”arXiv preprint arXiv:2411.14110, 2024

2024 arXiv
[30]

Is my data in your retrieval database? membership inference attacks against retrieval augmented generation,

M. Anderson, G. Amit, and A. Goldsteen, “Is my data in your retrieval database? membership inference attacks against retrieval augmented generation, ”arXiv preprint arXiv:2405.20446, 2024

2024 arXiv
[31]

Mask-based membership infer- ence attacks for retrieval-augmented generation,

M. Liu, S. Zhang, and C. Long, “Mask-based membership infer- ence attacks for retrieval-augmented generation, ” arXiv preprint arXiv:2410.20142, 2024

2024 arXiv
[32]

Generating is believing: Member- ship inference attacks against retrieval-augmented generation,

Y. Li, G. Liu, C. Wang, and Y. Yang, “Generating is believing: Member- ship inference attacks against retrieval-augmented generation, ”arXiv preprint arXiv:2406.19234, 2024

2024 arXiv
[33]

Badrag: Identify- ing vulnerabilities in retrieval augmented generation of large language models,

J. Xue, M. Zheng, Y. Hu, F. Liu, X. Chen, and Q. Lou, “Badrag: Identify- ing vulnerabilities in retrieval augmented generation of large language models, ”arXiv preprint arXiv:2406.00083, 2024

2024 arXiv
[34]

Frag: Toward federated vector database management for col- laborative and secure retrieval-augmented generation,

D. Zhao, “Frag: Toward federated vector database management for col- laborative and secure retrieval-augmented generation, ”arXiv preprint arXiv:2410.13272, 2024

2024 arXiv
[35]

Rag-wm: An efficient black-box watermarking approach for retrieval-augmented generation of large language models,

P. Lv, M. Sun, H. Wang, X. Wang, S. Zhang, Y. Chen, K. Chen, and L. Sun, “Rag-wm: An efficient black-box watermarking approach for retrieval-augmented generation of large language models, ”arXiv preprint arXiv:2501.05249, 2025

2025 arXiv
[36]

Rag with differential privacy,

N. Grislain, “Rag with differential privacy, ” arXiv preprint arXiv:2412.19291, 2024

2024 arXiv
[37]

A light- weight and attack-proof bidirectional blockchain paradigm for internet of things,

C. Xu, Y. Qu, T. H. Luan, P. W. Eklund, Y. Xiang, and L. Gao, “A light- weight and attack-proof bidirectional blockchain paradigm for internet of things, ”IEEE Internet of Things Journal , vol. 9, no. 6, pp. 4371–4384, 2021

2021
[38]

Error and attack tolerance of complex networks,

R. Albert, H. Jeong, and A.-L. Barabási, “Error and attack tolerance of complex networks, ”nature, vol. 406, no. 6794, pp. 378–382, 2000. Received 20 February 2007; revised 12 March 2009; accepted 5 June 2009

2000

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

A review on edge large language models: Design, execution, and applications,

Y. Zheng, Y. Chen, B. Qian, X. Shi, Y. Shu, and J. Chen, “A review on edge large language models: Design, execution, and applications, ” ACM Computing Surveys, 2024

work page 2024

[2] [2]

Mobile edge intelligence for large language models: A contemporary survey,

G. Qu, Q. Chen, W. Wei, Z. Lin, X. Chen, and K. Huang, “Mobile edge intelligence for large language models: A contemporary survey, ”IEEE Communications Surveys & Tutorials, 2025

work page 2025

[3] [3]

Pre-trained language models in biomedical domain: A systematic survey,

B. Wang, Q. Xie, J. Pei, Z. Chen, P. Tiwari, Z. Li, and J. Fu, “Pre-trained language models in biomedical domain: A systematic survey, ”ACM Computing Surveys, vol. 56, no. 3, pp. 1–52, 2023

work page 2023

[4] [4]

Vision language models in autonomous driving: A survey and outlook,

X. Zhou, M. Liu, E. Yurtsever, B. L. Zagar, W. Zimmer, H. Cao, and A. C. Knoll, “Vision language models in autonomous driving: A survey and outlook, ”IEEE Transactions on Intelligent Vehicles, 2024

work page 2024

[5] [5]

Survey of hallucination in natural language generation,

Z. Ji, N. Lee, R. Frieske, T. Yu, D. Su, Y. Xu, E. Ishii, Y. J. Bang, A. Madotto, and P. Fung, “Survey of hallucination in natural language generation, ” ACM computing surveys, vol. 55, no. 12, pp. 1–38, 2023

work page 2023

[6] [6]

Retrieval-augmented generation for large language models: A survey,

Y. Gao, Y. Xiong, X. Gao, K. Jia, J. Pan, Y. Bi, Y. Dai, J. Sun, and H. Wang, “Retrieval-augmented generation for large language models: A survey, ” arXiv preprint arXiv:2312.10997, 2023

arXiv 2023

[7] [7]

A survey on rag meeting llms: Towards retrieval-augmented large language models,

W. Fan, Y. Ding, L. Ning, S. Wang, H. Li, D. Yin, T.-S. Chua, and Q. Li, “A survey on rag meeting llms: Towards retrieval-augmented large language models, ” inProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , 2024, pp. 6491–6501

work page 2024

[8] [8]

Don’t forget private retrieval: distributed private similarity search for large language models,

G. Zyskind, T. South, and A. Pentland, “Don’t forget private retrieval: distributed private similarity search for large language models, ” inPro- ceedings of the Fifth Workshop on Privacy in Natural Language Process- ing. Bangkok, Thailand: Association for Computational Linguistics, Aug. 2024, pp. 7–19

work page 2024