RACORN: Robust Predicate-Agnostic Filtered Vector Search under Extreme Selectivity
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Computer Science. The day in computer science: software, systems and theory on arXiv.
Tom: We're looking at the paper: "RACORN: Robust Predicate-Agnostic Filtered Vector Search under Extreme Selectivity".
Jane: Here's the gist of the paper. Filtered Vector Search (FVS), which combines vector embedding similarity with structured metadata predicates, has emerged as a core requirement in RAG and production retrieval systems.
Tom: ACORN1, the representative In-filtering algorithm that reuses an existing HNSW index, substantially reduces latency at low selectivity but suffers connectivity instability below 5% selectivity and recall collapse below 1%.
Jane: We propose RACORN-1, an in-place extension of ACORN-1 that resolves this collapse via (i) Adaptive Search Fallback (ASF) — repurposing filter-failing nodes as transient bridges to detour around severed paths; bridge and two-hop candidate selection uses stride sampling for spatial diversity.
Tom: Across three 1M-scale and one 40M-scale dataset, RACORN-1 delivers approximately 9–26× latency reduction over HNSW in the sweet spot (1%–0.3%), and recovers ACORN-1’s recall collapse from 0.45–0.72 (1%) and 0.03–0.10 (0.3%) to 0.70–96 and 87–98 respectively.
Title and authors: Tom: So we're looking at this paper called RACORN: Robust Predicate-Agnostic Filtered Vector Search under Extreme Selectivity. Jane, what's the big picture here for a listener who doesn't know what a vector search is?
Jane: Well, basically, they’re talking about combining two ways to search data: using vector embeddings for similarity and using structured metadata predicates to narrow things down. Tom, the title tells us this paper is tackling the problem of keeping that filtering working when you have very few matching results.
Lu: Exactly. The authors are focused on an algorithm called ACORN1, which is a type of in-filtering algorithm that uses an existing HNSW index. They found that while ACORN1 makes searching faster when there are few matches, it starts having problems—specifically connectivity instability and recall collapse—when the selectivity drops below about five percent and the recall drops below one percent <ref:2607.00768#pg1>.
Meng: What does "recall collapse" mean for us in production? It means if we only get a small fraction of results, the system starts missing important data entirely, which is a huge problem when you're trying to find something specific.
Tom: Right. And the authors are proposing RACORN-one as an extension to ACORN1 that fixes that collapse by adding two main things: first, an Adaptive Search Fallback or ASF, and second, stride sampling for spatial diversity <ref:2607.00768#pg1>.
Jane: It sounds like they’re trying to make the search structure smarter so it doesn't just stop working when the number of matches gets really low.
Lu: That's right. They are repurposing those nodes that fail the predicate checks into transient bridges, essentially building detour paths around sections that are blocked <ref:2607.00768#pg1>.
The paper's summary: Tom: So, what’s the core mechanism of this RACORN-one <ref:2607.00768#pg1>? Jane, can you explain how this adaptive fallback works in plain English for someone who isn't deep into graph algorithms?
Jane: Okay. When the search runs and it hits a node where the filter condition fails—meaning it doesn't satisfy our metadata predicate—instead of just stopping or failing that path, RACORN-one treats that failed node as a temporary bridge <ref:2607.00768#pg1>.
Lu: That bridge count isn't fixed; it adapts. If the number of good candidates it finds in one hop and two hops is too low, the system grows that bridge count by the deficit to keep exploring <ref:2607.00768#pg1>.
Meng: So if we’re looking for something rare, instead of just giving up on a path, it keeps trying to find a way around the dead ends? That sounds like it could help us find those rare items.
Tom: It does. And they use stride sampling for selection when there are too many candidates—when the pool gets bigger than some target size T—to make sure they aren't just stuck on one neighbor <ref:2607.00768#pg1>.
Jane: That sampling part sounds like it keeps things diverse, which is important so you don't just get tunnel vision on one type of result. So, to summarize, RACORN-one’s whole goal is to maintain a search frontier whose size doesn't depend on how selective the search is <ref:2607.00768#pg1>.
Lu: That’s the structural fix they are aiming for. The reason ACORN1 had that recall collapse is because below a certain selectivity, even two-hop expansion doesn't find enough predicate-satisfying candidates to keep the search queue going <ref:2607.00768#pg3>.
The paper's improvements: Tom: Now let’s look at the actual numbers they got. Jane, what did they find when they tested RACORN-one against the original ACORN1 <ref:2607.00768#pg1>? We need concrete results here.
Jane: They showed that across three one million scale datasets and one forty million scale dataset, RACORN-one delivers about nine to twenty-six times fewer milliseconds of latency compared to HNSW in the sweet spot, which is when selectivity is between one percent and zero point three percent <ref:2607.00768#pg1>.
Meng: That’s a big speedup. Nine to twenty-six times faster means we can handle way more queries in the same time window, which is exactly what we need for production systems. But what about the recall numbers they recovered?
Lu: They showed that RACORN-one recovers ACORN1’s recall collapse significantly <ref:2607.00768#pg1>. For example, when selectivity was at one percent, ACORN1’s recall dropped to between zero point four five and zero point seven two, but RACORN-one brought that up to seventy percent or higher <ref:2607.00768#pg1>.
Tom: And for the lowest selectivity case, when it was at zero point three percent, ACORN1’s recall was down to maybe just three percent or ten percent, and RACORN-one got that up to eighty-seven or ninety eight percent <ref:2607.00768#pg1>.
Jane: That jump from maybe ten percent back up to nearly ninety eight is what really tells you it solved the recall collapse issue they were worried about with ACORN1 <ref:2607.00768#pg1>.
Lu: And they also introduced a variant, RACORN-one plus, which combines this with an Adaptive Exact Fallback or AEF when things get really extreme <ref:2607.00768#pg1>.
Conclusion: Tom: So, to wrap up, what’s the big takeaway here for us right now? Jane, can you summarize how RACORN-one and RACORN-one plus change how we think about filtered vector search systems <ref:2607.00768#pg1>?
Jane: It means that filter-first methods don't have to choose between speed and finding results. RACORN-one uses that bridge mechanism to keep the search going even when the data is sparse, andRACORN-one plus adds an exact search switch when things are really low on selectivity <ref:2607.00768#pg1>.
Lu: The paper proves that you can get a structural improvement in how the algorithm searches without sacrificing recall at the expense of latency in those challenging low-selectivity scenarios <ref:2607.00768#pg3>.
Meng: From an engineering standpoint, this means we have a better tool to deploy when we know our data might be very sparse, because it won't just break down under those conditions <ref:2607.00768#pg1>.
Tom: And they also gave us that bridge ratio knob you can tune between ACORN1 and HNSW behavior, which lets you trade latency for recall based on what your specific dataset needs <ref:2607.00768#pg1>.
Jane: It sounds like a very practical tool for production where you need reliability across different data distributions <ref:2607.00768#pg1>.
Lu: And RACORN-one plus is really powerful because it handles the extreme low selectivity regime, like zero point three percent on one million records, giving you recall one point zero with a speedup of twenty to seventy times <ref:2607.00768#pg1>.
Meng: That speedup at that level is huge. If we can switch to linear scan at that low percentage, the performance gains are substantial <ref:2607.00768#pg1>.
Tom: So, RACORN-one and RACORN-one plus give us a robust way to handle those tricky situations where the standard filter search breaks down <ref:2607.00768#pg1>.
Jane: That’s it for this paper on RACORN: Robust Predicate-Agnostic Filtered Vector Search under Extreme Selectivity. We’ll see what else is coming in the next episode.
Yoonseok Kim, Gyusik Choe
Naver Corporation
cs.DB, cs.IR
Submitted: 2026-07-01
Updated: 2026-11-06
Comments: 13 pages, 8 figures, 4 tables. Code: https://github.com/naver/racorn
Code: https://github.com/nmslib/hnswlib
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 85/100
The gist: Filtered Vector Search (FVS), which combines vector embedding similarity with structured metadata predicates, has emerged as a core requirement in RAG and production retrieval systems.
Key concepts
- Adaptive Search Fallback (ASF)
- This mechanism repurposes nodes that fail the filter search as 'transient bridges' to bypass broken paths in the graph. The number of these bridges grows if there aren't enough good candidates, ensuring the search frontier remains active and doesn't stall when selectivity is low.
- Stride Sampling
- Used during candidate selection, stride sampling ensures spatial diversity by extracting candidates uniformly at a specific interval (sigma) from the pool of filter-failing nodes. This prevents the search from focusing too heavily on just one neighbor, promoting a more radial exploration of the search space.
- Bridge Ratio Knob (BR)
- This adjustable parameter allows users to interpolate between different behaviors: ACORN-1's high latency advantage and HNSW's recall recovery. Tuning this knob lets operators trade search speed for better recall based on how selective their data is.
Terminology
Summary
Filtered Vector Search (FVS), which combines vector embedding similarity with structured metadata predicates, has emerged as a core requirement in RAG and production retrieval systems. p. 0 ACORN1, the representative In-filtering algorithm that reuses an existing HNSW index, substantially reduces latency at low selectivity but suffers connectivity instability below 5% selectivity and recall collapse below 1%. p. 0 We propose RACORN-1, an in-place extension of ACORN-1 that resolves this collapse via (i) Adaptive Search Fallback (ASF) — repurposing filter-failing nodes as transient bridges to detour around severed paths; bridge and two-hop candidate selection uses stride sampling for spatial diversity. p. 0 Across three 1M-scale and one 40M-scale dataset, RACORN-1 delivers approximately 9–26× latency reduction over HNSW in the sweet spot (1%–0. p. 03%), and recovers ACORN-1’s recall collapse from 0. p. 045–0.72 (1%) and 0.03–0.10 (0.3%) to 0.70–96 and 87–98 respectively. p. 0 For the extreme-low-selectivity regime where linear scan can outperform graph search, we combine RACORN-1 with (ii) Adaptive Exact Fallback (AEF) in a variant RACORN-1+, achieving recall 1. p. 000 with 20–75× speedup at 1M ≤ 0. p. 01% and 40M 0. p. 001%.
Pruning redundant reasoning tokens
The paper introduces RACORN-1 to resolve the low-selectivity recall collapse observed in ACORN-1 by repurposing filter-failing nodes as transient bridges and employing stride sampling for spatial diversity. This mechanism is designed to maintain a search frontier whose size is independent of selectivity, addressing the structural cause of the recall collapse which stems from the shrinking count of predicate-satisfying candidates below a critical selectivity and. p. 0
How the training signal is built
RACORN-1's core mechanism involves Adaptive Search Fallback (ASF), where filter-failing nodes are admitted into the search queue as transient bridges when one-hop and two-hop predicate-satisfying candidates are insufficient. The bridge count adapts to both the unvisited one-hop count and the two-hop predicate-satisfying count, growing by the deficit if candidates are insufficient. p. 2 Stride sampling is used to reinforce spatial diversity when candidate pools exceed a target size T, extracting candidates uniformly at stride sigma to ensure radial diversity around the current node. p. 3
Where the gains come from
The analysis in section 4.1 and 4.5 develops an s-independent speedup through a region-wise cost analysis and a formal justification of the s-independent expansion-size lower bound under a random graph model. p. 0 Theorem 1 establishes that the expected expansion size E[expand(u)] ≥ M · b for s < b/M, proving that RACORN-1’s traversal frontier does not stall while ACORN-1’s does. p. 5 Empirical validation confirms this structural difference through the ACORN-1 vs RACORN-1 recall delta in the baseline setting. p. 5
Structural Trade-offs and Regimes
The paper contrasts filter-first methods with distance-first HNSW by showing that while filter-first methods reduce latency at low selectivity, they suffer a structural recall trade-off relative to distance-first HNSW. RACORN-1 mitigates this trade-off via ASF, and the bridge ratio knob acts as a continuous interpolation between ACORN1 and HNSW In-filtering behaviors. p. 3 Furthermore, RACORN-1+ incorporates Adaptive Exact Fallback (AEF) to switch to an exact search when the running predicate-pass ratio falls below a threshold, ensuring recall 1. p. 00 with substantial speedup in the extreme-low-selectivity regime. p. 0
Operational Robustness and Correlation
The work investigates the effect of query-filter correlation (Pos/Neg) on the ACORN family, showing that under negative correlation, RACORN-1’s ASF acts as a safety net preserving recall at 0.82–0.98 with a 5–9× latency advantage over HNSW. p. 0 The BR sensitivity analysis reveals that under negative correlation, BR↓ emerges as the efficient frontier for latency efficiency. p. 9 Finally, RACORN-1+ demonstrates that the AEF exact switch can activate at selectivities as low as 0. p. 73% on 1M datasets to guarantee recall 1. p. 20 with up to 28× speedup. p. 0
Conclusion
Together, RACORN-1 and RACORN-1+ form an ACORN1-compatible mechanism robust to both extreme-low-selectivity and adversarial query–filter correlation. The newly introduced bridge ratio knob enables additional recall in the ASF-activation region (≤ 5%), supplying +0. p. 3 p. 312 to +0.19 recall gain on datasets with lower baseline recall. p. 9 p. 9 RACORN-1+ handles the extreme-low-selectivity regime where graph search underperforms linear scan — at ≤ 0. p. 0 p. 01% on 1M, recall 1. p. 0 p. 000 with 20–75× speedup. p. 0 p. 0 Together, RACORN-1 and RACORN-1+ defend recall across the extreme-low-selectivity and negative-correlation regimes where ACORN-1 fails. p. 0 p. 0
References
[1] Martin Aumüller, Erik Bernhardsson, and Alexander Faithfull. 2020. ANNBenchmarks: A Benchmarking Tool for Approximate Nearest Neighbor Algorithms. p. 0 p. 0
[2] Qi Chen, Bing Zhao, Haidong Wang, Mingqin Li, Chuanjie Liu, Zengzhong Li, Mao Yang, and Jingdong Wang. 2021. SPANN: Highly-Efficient Billion-Scale Approximate Nearest Neighborhood Search. p. 0 p. 0
[3] Matthijs Douze, Alexandr Guzhva, Chengqi Deng, Jeff Johnson, Gergely Szilvasy, Pierre-Emmanuel Mazaré, Maria Lomeli, Lucas Hosseini, and Hervé Jégou. 2024. The Faiss Library. p. 0 p. 0
[4] Elastic. [n.d.]. Filtered HNSW kNN Search. p. 0 p. 0
[5] Cong Fu, Chao Xiang, Changxu Wang, and Deng Cai. 2019. Fast Approximate Nearest Neighbor Search with the Navigating Spreading-out Graph. p. 0 p. 0
[6] Siddharth Gollapudi, Neel Karia, Varun Sivashankar, Ravishankar Krishnaswamy, Nikit Begwani, Swapnil Raz, Yiyong Lin, Yin Zhang, Neelam Mahapatro, Premkumar Srinivasan, Amit Singh and Harsha Vardhan Simhadri. 2023. FilteredDiskANN: Graph Algorithms for Approximate Nearest Neighbor Search with Filters. p. 0 p. 0
[7] Ruiqi Guo, Philip Sun, Erik Lindgren, Quan Geng, David Simcha, Felix Chern, and Sanjiv Kumar. 2020. Accelerating Large-Scale Inference with Anisotropic Vector Quantization. p. 6 p. 6
[8] Hervé Jégou, Matthijs Douze, and Cordelia Schmid. 2011. Product Quantization for Nearest Neighbor Search. p. 2 p. 2
[9] Patrick Lewis, Ethan Perez, Aleksandra Piktus, et al. 2020. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. p. 0 p. 0
[10] J. MacQueen. 1967. Some Methods for Classification and Analysis of Multivariate Observations. p. 0 p. 0
[11] Yu. A. Malkov et al. [n.d.]. hnswlib: Header-only C++/Python Library for Fast Approximate Nearest Neighbors. p. 0 p. 0
[12] Yu. A. Malkov and D. A. Yashunin. 2018. Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs. p. 1 p. 1
[13] James Jie Pan, Jianguo Wang, and Guoliang Li. 2024. Survey of Vector Database Management Systems. p. 0 p. 0
[14] Liana Patel, Peter Kassoumeh, Carlos Guibas, et al. 2024. ACORN: Performant and Predicate-Agnostic Search Over Vector Embeddings and Structured Data. p. 0 p. 0
[15] Qdrant. p. 2 p. 2 [n.d.]. ACORN Algorithm in Qdrant 1. p. 2 p. 216.
[16] P. Griffiths Selinger, M. M. Astrahan, D. D. Chamberlin, R. A. Lorie, and T. G. Price. 1979.
[17] Harsha Vardhan Simhadri, George Williams, Martin Aumüller et al. 2022. Results of the NeurIPS 2021 Challenge on Billion-Scale Approximate Nearest Neighbor Search. p. 0 p. 0
[18] Suhas Jayaram Subramanya, Devvrit, Harsha Vardhan Simhadri et al. 2019. DiskANN: Fast Accurate Billion-Point Nearest Neighbor Search on a Single Node. p. 0 p. 0
[19] Philip Sun, David Simcha, Dave Dopson, Ruiqi Guo, and Sanjiv Kumar. 2023. SOAR: Improved Indexing for Approximate Nearest Neighbor Search. p. 0 p. 0
[20] Unum Cloud. [n.d.]. USearch: Smaller and Faster Single-File Vector Search Engine. p. 0 p. 0
[21] Vespa. p. 2 p. 2 [n.d.]. Additions to HNSW. p. 0 p. 0
[22] Jianguo Wang et al. 2021. Milvus: A Purpose-Built Vector Data Management System. p. 2 p. 2
[23] Mengzhao Wang et al. 2021. Navigable Proximity Graph-Driven Native Hybrid Queries with Structured and Unstructured Constraints. p. 0 p. 0
[24] Mengzhao Wang et al. 2021. A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search. p. 0 p. 0
[25] Weaviate. p. 2 p. 2 2024. Speed-up Filtered Vector Search. p. 0 p. 0
[26] Chuangxian Wei et al. 2020. AnalyticDB-V: A Hybrid Analytical Engine Towards Query Fusion for Structured and Unstructured Data. p. 0 p. 0
Together, RACORN-1 and RACORN-1+ defend recall across the extreme-low-selectivity and negative-correlation regimes where ACORN-1 fails. The newly introduced bridge ratio knob enables additional recall in the ASF-activation region (≤ 5%), supplying +0.12 to +0.19 recall gain on datasets with lower baseline recall. RACORN-1+ handles the extreme-low-selectivity regime where graph search underperforms linear scan — at ≤ 0.1% on 1M, recall 1.00 with 20–75× speedup. Together, RACORN-1 and RACORN-1+ defend recall across the extreme-low-selectivity and negative-correlation regimes where ACORN-1 fails.
[1] Martin Aumüller, Erik Bernhardsson, and Alexander Faithfull. 2020. ANNBenchmarks: A Benchmarking Tool for Approximate Nearest Neighbor Algorithms.
[2] Qi Chen, Bing Zhao, Haidong Wang, et al. 2021. SPANN: Highly-Efficient Billion-Scale Approximate Nearest Neighborhood Search.
[3] Matthijs Douze et al. 2024. The Faiss Library.
[4] Elastic. [n.d.]. Filtered HNSW kNN Search.
[5] Cong Fu et al. 2019. Fast Approximate Nearest Neighbor Search with the Navigating Spreading-out Graph.
[6] Siddharth Gollapudi et al. 2023. FilteredDiskANN: Graph Algorithms for Approximate Nearest Neighbor Search with Filters.
[7] Ruiqi Guo et al. 2020. Accelerating Large-Scale Inference with Anisotropic Vector Quantization.
[8] Hervé Jégou et al. 2011. Product Quantization for Nearest Neighbor Search.
[9] Patrick Lewis et al. 2020. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks.
[10] J. MacQueen. 1967. Some Methods for Classification and Analysis of Multivariate Observations.
[11] Yu. A. Malkov et al. [n.d.]. hnswlib: Header-only C++/Python Library for Fast Approximate Nearest Neighbors.
[12] Yu. A. Malkov and D. A. Yashunin 2018. Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs.
[13] James Jie Pan et al. 2024. Survey of Vector Database Management Systems.
[14] Liana Patel et al. 2024. ACORN: Performant and Predicate-Agnostic Search Over Vector Embeddings and Structured Data.
[15] Qdrant. [n.d.]. ACORN Algorithm in Qdrant 1.16.
[16] P. Griffiths Selinger et al. 1979.
[17] Harsha Vardhan Simhadri et al. 2022. Results of the NeurIPS 2021 Challenge on Billion-Scale Approximate Nearest Neighbor Search.
[18] Suhas Jayaram Subramanya et al. 2019. DiskANN: Fast Accurate Billion-Point Nearest Neighbor Search on a Single Node.
[19] Philip Sun et al. 2023. SOAR: Improved Indexing for Approximate Nearest Neighbor Search.
[20] Unum Cloud. [n.d.]. USearch: Smaller and Faster Single-File Vector Search Engine.
[21] Vespa. [n.d.]. Additions to HNSW.
[22] Jianguo Wang et al. 2021. Milvus: A Purpose-Built Vector Data Management System.
[23] Mengzhao Wang et al. 2021. Navigable Proximity Graph-Driven Native Hybrid Queries with Structured and Unstructured Constraints.
[24] Mengzhao Wang et al. 2021. A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search.
[25] Weaviate. 2024. Speed-up Filtered Vector Search.
[26] Chuangxian Wei et al. 2020. AnalyticDB-V: A Hybrid Analytical Engine Towards Query Fusion for Structured and Unstructured Data.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Adaptive Recall-Preserving Speedup for Low-Selectivity Filtered Vector Search (RACORN-1) resolves recall collapse by reinterpreting filter-failing nodes as
transient bridges to detour around severed paths,
ensuring the traversal frontier does not stall even when one-hop and two-hop predicate-satisfying candidates are insufficient. -
The RACORN-1+ variant with Adaptive Exact Fallback (AEF) enables the system to switch automatically to an Exact Search when the running predicate-pass ratio falls below a threshold, achieving recall 1.00 with substantial speedup in extreme low-selectivity regimes like 0.3% on 1M datasets.
-
Stride sampling is applied during bridge and two-hop candidate selection to reinforce spatial diversity, ensuring that the search does not concentrate on a particular one-hop neighbor by extracting candidates uniformly at stride sigma from the pool of filter-failing nodes when the pool exceeds a target size T.
-
The system can dynamically adjust a bridge ratio knob (BR) between zero and M, acting as an
interpolation knob
to tune behavior between ACORN-1 (maximum latency advantage) and HNSW In-filtering (recall recovery), allowing operators to trade latency for recall based on the specific dataset's baseline recall. -
The RACORN-1+ system can determine the Adaptive Exact Fallback threshold dynamically by scaling it proportionally to the search thread count (EFs) and the data scale N, ensuring optimal activation timing for linear scan switch based on cost-benefit analysis between graph search and brute-force scan.
Sources
- ANN-Benchmarks: A Benchmarking Tool for Approximate Nearest Neighbor Algorithms
- SPANN: Highly-efficient Billion-scale Approximate Nearest Neighbor Search
- The Faiss library
- Fast Approximate Nearest Neighbor Search With The Navigating Spreading-out Graph
- Accelerating Large-Scale Inference with Anisotropic Vector Quantization
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
- Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs
- ACORN: Performant and Predicate-Agnostic Search Over Vector Embeddings and Structured Data
- SOAR: Improved Indexing for Approximate Nearest Neighbor Search
- Navigable Proximity Graph-Driven Native Hybrid Queries with Structured and Unstructured Constraints
- A Comprehensive Survey and Experimental Comparison of Graph-Based Approximate Nearest Neighbor Search
Related papers
- Beyond Overlap: Hierarchical Graph Abstraction for Out-of-Core Vector Search
- EMA: Approximate Nearest Neighbor Search with General Attribute Filtering and Dynamic Updates
- Open Biomedical Knowledge Graphs at Scale: Construction, Federation, and AI Agent Access with Samyama Graph Database
- Samyama: A Unified Graph-Vector Database with In-Database Optimization, Agentic Enrichment, and Hardware Acceleration
- GenTUS: Table Union Search via Constrained Generation of Semantic Table Identifiers
- Common-Neighbor-Count-Based Representative Possible World Finding on Uncertain Graphs