pith. sign in

arxiv: 2606.04610 · v1 · pith:P52I7TCYnew · submitted 2026-06-03 · 💻 cs.DB

Selectivity Estimation for Semantic Filters on Image Data

Matthias Urban , Vu Huy Nguyen , Gabriele Sanmartino , Paolo Papotti , Carsten Binnig This is my paper

Pith reviewed 2026-06-28 04:04 UTC · model grok-4.3

classification 💻 cs.DB
keywords selectivity estimationsemantic filtersimage dataembedding spacesquery optimizationsemantic histogramsvision language models
0
0 comments X

The pith

Semantic Histograms estimate selectivity of semantic image filters by treating them as implicit range queries in shared embedding spaces.

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

The paper shows how to estimate the fraction of images that match a semantic filter without running online samples during query time. It observes that every semantic filter corresponds to some range of points in an embedding space, with more general filters covering wider ranges and more specific ones covering narrower ranges. Two methods are developed to gauge the width of these implicit ranges, and their ensemble delivers the strongest estimates. When used in place of sampling, the approach reduces the combined time for optimizing and running semantic queries by as much as 86 percent.

Core claim

Semantic Histograms supply selectivity estimates for semantic filters on image data by quantifying the specificity of the implicit ranges those filters induce in shared embedding spaces. Two distinct estimation procedures are introduced, and an ensemble of the two produces the most accurate results, allowing query optimizers to bypass the latency of online sample-based profiling.

What carries the argument

Semantic Histograms, which model semantic filters as implicit range queries whose range widths are estimated from embedding-space geometry.

If this is right

  • Query optimizers can determine execution orders for semantic filters without incurring sampling latency.
  • Low-selectivity semantic filters become feasible to handle in interactive database workloads.
  • An ensemble of the two range-width estimators outperforms either method used alone.
  • Semantic data systems that embed LLMs and VLMs can reduce overall query runtime overhead.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same embedding-space approach could be applied to semantic filters over text or video once suitable shared spaces exist.
  • Precomputed histograms could be maintained incrementally as new images are added to a collection.
  • Hybrid estimators that fall back to limited sampling only on uncertain predictions might further improve accuracy.

Load-bearing premise

Semantic filters correspond to well-behaved implicit ranges in the embedding space whose sizes can be estimated accurately without inspecting actual data samples or model internals at query time.

What would settle it

Measure the actual fraction of images that satisfy each test filter on a held-out image collection and compare those ground-truth selectivities against the estimates produced by Semantic Histograms.

Figures

Figures reproduced from arXiv: 2606.04610 by Carsten Binnig, Gabriele Sanmartino, Matthias Urban, Paolo Papotti, Vu Huy Nguyen.

Figure 1
Figure 1. Figure 1: (Left) The semantic filter selects those pictures that depict a [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (Top) Offline, we feed a selected number of images [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Semantic Histograms are much faster than sampling, while being very accurate. We plot the Q-Error (Y-axis) of [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Using Semantic Histograms for query optimization leads to faster end-to-end runtime than sampling. We generate [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
read the original abstract

Semantic data systems integrate Large Language Models (LLMs) and Vision-Language Models (VLMs) directly into database query execution, enabling expressive queries on multi-modal data. However, optimizing these queries requires accurate selectivity estimates to determine the most efficient operator execution order. Contemporary systems rely on online sample-based profiling, a process that incurs severe latency overheads and struggles with low-selectivity queries. In this paper, we introduce Semantic Histograms, a novel selectivity estimator for semantic filters on image data that leverages shared embedding spaces to bypass traditional profiling. We realize that all semantic filters are implicit range queries, as they match a range of different images. Some filter predicates are more general, yielding a wide range, while others are more specific, yielding a smaller range. To address the challenge of implicit ranges, we propose two approaches to estimate the queries' specificity, with an ensemble of the two performing best. The evaluation shows that Semantic Histograms can reduce the end-to-end runtime overhead of query optimization and execution by up to 86%.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper introduces Semantic Histograms as a selectivity estimator for semantic filters on image data. It models every semantic filter as an implicit range query in a shared embedding space, proposes two (unnamed) methods to estimate query specificity without online sampling or data access, reports that an ensemble of the two performs best, and claims this reduces end-to-end runtime overhead of query optimization and execution by up to 86%.

Significance. If the estimators can be shown to be both cheap and sufficiently accurate (especially for low-selectivity predicates), the approach could meaningfully improve query planning in semantic data systems that integrate VLMs, by eliminating the latency of sample-based profiling.

major comments (2)
  1. [Abstract] Abstract: the central claim of an 86% end-to-end overhead reduction is stated without any description of the datasets, baselines, error metrics, implementation of the two specificity estimators, or how the ensemble is constructed. These details are load-bearing for assessing whether the estimates are accurate enough to avoid worse query plans.
  2. [Abstract] Abstract: the load-bearing assumption that semantic filters map to well-behaved, estimable implicit ranges in embedding space (whose specificity can be computed without inspecting data samples or model internals) is asserted but not supported by any definition of the two methods or any selectivity-error numbers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the abstract. We agree that the abstract is currently too concise and does not provide sufficient context for the central claims. We will revise the abstract to include high-level descriptions of the methods, ensemble construction, datasets, baselines, and key error metrics while preserving its brevity.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim of an 86% end-to-end overhead reduction is stated without any description of the datasets, baselines, error metrics, implementation of the two specificity estimators, or how the ensemble is constructed. These details are load-bearing for assessing whether the estimates are accurate enough to avoid worse query plans.

    Authors: We agree that the abstract would be strengthened by including these details. In the revised manuscript we will expand the abstract to briefly name and characterize the two specificity estimators, describe how the ensemble is formed, list the primary datasets and baselines, and report the selectivity error metrics that underpin the 86% end-to-end overhead reduction. These additions will allow readers to evaluate the claim without immediately consulting the body of the paper. revision: yes

  2. Referee: [Abstract] Abstract: the load-bearing assumption that semantic filters map to well-behaved, estimable implicit ranges in embedding space (whose specificity can be computed without inspecting data samples or model internals) is asserted but not supported by any definition of the two methods or any selectivity-error numbers.

    Authors: We acknowledge the observation. The body of the manuscript defines the two methods (which operate solely on embedding-space statistics without data samples or model internals) and presents selectivity-error results. To address the concern directly in the abstract, the revision will incorporate a concise characterization of the methods together with representative selectivity-error figures, thereby supporting the implicit-range assumption at the abstract level. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained via empirical proposal

full rationale

The paper introduces Semantic Histograms as a new selectivity estimator based on the observation that semantic filters act as implicit range queries in embedding spaces, with two proposed estimation methods and their ensemble. No equations, fitted parameters, or derivations are presented that reduce the claimed selectivity estimates or 86% overhead reduction to self-definitional constructs, renamed known results, or load-bearing self-citations. The approach is positioned as bypassing profiling through direct use of embeddings, with accuracy evaluated empirically rather than derived by construction from inputs. This is the common case of an independent proposal without circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The central claim rests on the unelaborated premise that embedding-space range semantics hold for arbitrary semantic predicates.

pith-pipeline@v0.9.1-grok · 5711 in / 1282 out tokens · 22153 ms · 2026-06-28T04:04:12.314105+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SemCEB: A Cardinality Estimation Benchmark for Semantic Operators

    cs.DB 2026-06 unverdicted novelty 7.0

    SemCEB is the first benchmark for cardinality estimation over semantic operators, evaluating sampling methods and Semantic Histograms on accuracy, cost, latency, and memory using 102 queries on a real-world dataset.

Reference graph

Works this paper leans on

41 extracted references · 23 canonical work pages · cited by 1 Pith paper · 5 internal anchors

  1. [1]

    Paritosh Aggarwal, Bowei Chen, Anupam Datta, Benjamin Han, Boxin Jiang, Nitish Jindal, Zihan Li, Aaron Lin, Pawel Liskowski, Jay Tayade, Dimitris Tsirogiannis, Nathan Wiegand, and Weicheng Zhao. 2025. Cortex AISQL: A Production SQL Engine for Unstructured Data.CoRRabs/2511.07663 (2025). https://doi.org/10.48550/ARXIV.2511.07663 arXiv:2511.07663

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2511.07663 2025

  2. [2]

    Shah, Benjamin Sowell, Dan Tecuci, Vinayak Thapliyal, and Matt Welsh

    Eric Anderson, Jonathan Fritz, Austin Lee, Bohou Li, Mark Lindblad, Henry Lindeman, Alex Meyer, Parth Parmar, Tanvi Ranade, Mehul A. Shah, Benjamin Sowell, Dan Tecuci, Vinayak Thapliyal, and Matt Welsh. 2025. The Design of an LLM-powered Unstructured Analytics System. In15th Conference on Innovative Data Systems Research, CIDR 2025, Amsterdam, The Netherl...

    2025

  3. [3]

    Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, Humen Zhong, Yuanzhi Zhu, Ming-Hsuan Yang, Zhaohai Li, Jianqiang Wan, Pengfei Wang, Wei Ding, Zheren Fu, Yiheng Xu, Jiabo Ye, Xi Zhang, Tianbao Xie, Zesen Cheng, Hang Zhang, Zhibo Yang, Haiyang Xu, and Junyang Lin. 2025. Qwen2.5-VL Technical ...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2502.13923 2025

  4. [4]

    Shu Chen, Deepti Raghavan, and Ugur Çetintemel. 2025. Continuous Prompts: LLM-Augmented Pipeline Processing over Unstructured Streams. CoRRabs/2512.03389 (2025). https://doi.org/10.48550/ARXIV.2512.03389 arXiv:2512.03389

    work page doi:10.48550/arxiv.2512.03389 2025

  5. [5]

    100x Cost & Latency Reduction: Performance Analysis of AI Query Approximation using Lightweight Proxy Models

    Yeounoh Chung, Rushabh Desai, Jian He, Yu Xiao, Thibaud Hottelier, Yves- Laurent Kom Samo, Pushkar Khadilkar, Xianshun Chen, Sam Idicula, Fatma Özcan, Alon Y. Halevy, and Yannis Papakonstantinou. 2026. 100x Cost & Latency Reduction: Performance Analysis of AI Query Approximation using Lightweight Proxy Models.CoRRabs/2603.15970 (2026). https://doi.org/10....

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv 2026

  6. [6]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Im- ageNet: A large-scale hierarchical image database. In2009 IEEE Computer So- ciety Conference on Computer Vision and Pattern Recognition (CVPR 2009), 20- 25 June 2009, Miami, Florida, USA. IEEE Computer Society, 248–255. https: //doi.org/10.1109/CVPR.2009.5206848

    work page doi:10.1109/cvpr.2009.5206848 2009

  7. [7]

    Alessio Devoto, Maximilian Jeblick, and Simon Jégou. 2025. Expected Attention: KV Cache Compression by Estimating Attention from Future Queries Distribu- tion.CoRRabs/2510.00636 (2025). https://doi.org/10.48550/ARXIV.2510.00636 arXiv:2510.00636

    work page doi:10.48550/arxiv.2510.00636 2025

  8. [8]

    Anas Dorbani, Sunny Yasser, Jimmy Lin, and Amine Mhedhbi. 2025. Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB.Proc. VLDB Endow.18, 12 (2025), 5415–5418. https://doi.org/10.14778/3750601.3750685

    work page doi:10.14778/3750601.3750685 2025

  9. [9]

    Uélison Jean Lopes dos Santos, Alessandro Ferri, Szilard Nistor, Riccardo Tom- masini, Carsten Binnig, and Manisha Luthra. 2026. Towards Multimodal Stream Processing Systems. InProceedings 29th International Conference on Extending Database Technology, EDBT 2026, Tampere, Finland, March 24-27, 2026, Wolf- gang Lehner, Vanessa Braganholo, Kostas Stefanidis...

    work page doi:10.48786/edbt.2026.51 2026

  10. [10]

    Philippe Flajolet, Éric Fusy, Olivier Gandouet, and Frédéric Meunier. 2007. Hyper- loglog: the analysis of a near-optimal cardinality estimation algorithm.Discrete mathematics & theoretical computer scienceProceedings (2007)

    2007

  11. [11]

    Benjamin Hilprecht, Andreas Schmidt, Moritz Kulessa, Alejandro Molina, Kris- tian Kersting, and Carsten Binnig. 2019. DeepDB: Learn from Data, not from Queries!CoRRabs/1909.00607 (2019). arXiv:1909.00607 http://arxiv.org/abs/ 1909.00607

    arXiv 2019

  12. [12]

    Addison Howard, Eunbyung Park, and Wendy Kan. 2018. ImageNet Ob- ject Localization Challenge. https://kaggle.com/competitions/imagenet-object- localization-challenge. Kaggle

    2018

  13. [13]

    Saehan Jo and Immanuel Trummer. 2024. ThalamusDB: Approximate Query Processing on Multi-Modal Data.Proc. ACM Manag. Data2, 3 (2024), 186. https: //doi.org/10.1145/3654989

    work page doi:10.1145/3654989 2024

  14. [14]

    Andreas Kipf, Thomas Kipf, Bernhard Radke, Viktor Leis, Peter Boncz, and Al- fons Kemper. 2019. Learned Cardinalities: Estimating Correlated Joins with Deep Learning. In9th Biennial Conference on Innovative Data Systems Re- search, CIDR 2019, Asilomar, CA, USA, January 13-16, 2019, Online Proceedings. www.cidrdb.org. https://vldb.org/cidrdb/2019/learned-c...

    2019

  15. [15]

    Jiale Lao, Andreas Zimmerer, Olga Ovcharenko, Tianji Cong, Matthew Russo, Gerardo Vitagliano, Michael Cochez, Fatma Özcan, Gautam Gupta, Thibaud Hot- telier, H. V. Jagadish, Kris Kissel, Sebastian Schelter, Andreas Kipf, and Immanuel Trummer. 2025. SemBench: A Benchmark for Semantic Query Processing En- gines.CoRRabs/2511.01716 (2025). https://doi.org/10....

    work page doi:10.48550/arxiv.2511.01716 2025

  16. [16]

    Evergreen: Efficient Claim Verification for Semantic Aggregates

    Alexander W. Lee, Benjamin Han, Shayak Sen, Samuel Yeom, Ugur Çetintemel, and Anupam Datta. 2026. Evergreen: Efficient Claim Verification for Semantic Aggregates.CoRRabs/2604.26180 (2026). https://doi.org/10.48550/ARXIV.2604. 26180 arXiv:2604.26180

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2604 2026

  17. [17]

    Bo Li, Kaichen Zhang, Hao Zhang, Dong Guo, Renrui Zhang, Feng Li, Yuanhan Zhang, Ziwei Liu, and Chunyuan Li. 2024. LLaVA-NeXT: Stronger LLMs Super- charge Multimodal Capabilities in the Wild. https://llava-vl.github.io/blog/2024- 05-10-llava-next-stronger-llms/

    2024

  18. [18]

    Zequn Li, Yuanhao Zhong, Chengliang Chai, Zhaoze Sun, Yuhao Deng, Ye Yuan, Guoren Wang, and Lei Cao. 2025. DocDB: A Database for Unstructured Document Analysis.Proc. VLDB Endow.18, 12 (2025), 5387–5390. https://doi.org/10.14778/ 3750601.3750678

    arXiv 2025

  19. [19]

    Parameswaran, and Eugene Wu

    Yiming Lin, Madelon Hulsebos, Ruiying Ma, Shreya Shankar, Sepanta Zeighami, Aditya G. Parameswaran, and Eugene Wu. 2024. Towards Accurate and Efficient Document Analytics with Large Language Models.CoRRabs/2405.04674 (2024). https://doi.org/10.48550/ARXIV.2405.04674 arXiv:2405.04674

    work page doi:10.48550/arxiv.2405.04674 2024

  20. [20]

    Cafarella, Lei Cao, Peter Baile Chen, Zui Chen, Michael J

    Chunwei Liu, Matthew Russo, Michael J. Cafarella, Lei Cao, Peter Baile Chen, Zui Chen, Michael J. Franklin, Tim Kraska, Samuel Madden, Rana Shahout, and Gerardo Vitagliano. 2025. Palimpzest: Optimizing AI-Powered Analyt- ics with Declarative Query Processing. In15th Conference on Innovative Data Systems Research, CIDR 2025, Amsterdam, The Netherlands, Jan...

    2025

  21. [21]

    https://vldb.org/cidrdb/2025/palimpzest-optimizing-ai- powered-analytics-with-declarative-query-processing.html

    www.cidrdb.org. https://vldb.org/cidrdb/2025/palimpzest-optimizing-ai- powered-analytics-with-declarative-query-processing.html

    2025

  22. [22]

    George A. Miller. 1995. WordNet: A Lexical Database for English.Commun. ACM38, 11 (1995), 39–41. https://doi.org/10.1145/219717.219748

    work page doi:10.1145/219717.219748 1995

  23. [23]

    Pan, Harshit Gupta, Parth Asawa, Carlos Guestrin, and Matei Zaharia

    Liana Patel, Siddharth Jha, Melissa Z. Pan, Harshit Gupta, Parth Asawa, Carlos Guestrin, and Matei Zaharia. 2025. Semantic Operators and Their Optimization: Towards AI-Based Data Analytics with Accuracy Guarantees.Proc. VLDB Endow. 18, 11 (2025), 4171–4184. https://doi.org/10.14778/3749646.3749685

    work page doi:10.14778/3749646.3749685 2025

  24. [24]

    Ioannidis, Peter J

    Viswanath Poosala, Yannis E. Ioannidis, Peter J. Haas, and Eugene J. Shekita

  25. [25]

    In Proceedings of the 1996 ACM SIGMOD International Conference on Management of Data, Montreal, Quebec, Canada, June 4-6, 1996, H

    Improved Histograms for Selectivity Estimation of Range Predicates. In Proceedings of the 1996 ACM SIGMOD International Conference on Management of Data, Montreal, Quebec, Canada, June 4-6, 1996, H. V. Jagadish and Inderpal Singh Mumick (Eds.). ACM Press, 294–305. https://doi.org/10.1145/233269.233342

    work page doi:10.1145/233269.233342 1996

  26. [26]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. 2021. Learning Transferable Visual Mod- els From Natural Language Supervision. InProceedings of the 38th International Conference on Machine Learning, ICML 2021, 18-24 Jul...

    2021

  27. [27]

    Cafarella, Tim Kraska, and Samuel Madden

    Matthew Russo, Chunwei Liu, Sivaprasad Sudhir, Gerardo Vitagliano, Michael J. Cafarella, Tim Kraska, and Samuel Madden. 2026. Abacus: A Cost-Based Opti- mizer for Semantic Operator Systems.Proc. VLDB Endow.19, 5 (2026), 1060–1073. https://www.vldb.org/pvldb/vol19/p1060-russo.pdf

    2026

  28. [28]

    Gabriele Sanmartino, Matthias Urban, Paolo Papotti, and Carsten Binnig

  29. [29]

    CoRRabs/2602.04430 (2026)

    The Stretto Execution Engine for LLM-Augmented Data Systems. CoRRabs/2602.04430 (2026). https://doi.org/10.48550/ARXIV.2602.04430 arXiv:2602.04430

    work page doi:10.48550/arxiv.2602.04430 2026

  30. [30]

    Dario Satriani, Enzo Veltri, Donatello Santoro, Sara Rosato, Simone Varriale, and Paolo Papotti. 2025. Logical and Physical Optimizations for SQL Query Execution over Large Language Models.Proc. ACM Manag. Data3, 3 (2025), 181:1–181:28. https://doi.org/10.1145/3725411

    work page doi:10.1145/3725411 2025

  31. [31]

    Selinger, Morton M

    Patricia G. Selinger, Morton M. Astrahan, Donald D. Chamberlin, Raymond A. Lorie, and Thomas G. Price. 1979. Access Path Selection in a Relational Database Management System. InProceedings of the 1979 ACM SIGMOD International Conference on Management of Data, Boston, Massachusetts, USA, May 30 - June 1, Philip A. Bernstein (Ed.). ACM, 23–34. https://doi.o...

    work page doi:10.1145/582095.582099 1979

  32. [32]

    Parameswaran, and Eugene Wu

    Shreya Shankar, Tristan Chambers, Tarak Shah, Aditya G. Parameswaran, and Eugene Wu. 2025. DocETL: Agentic Query Rewriting and Evaluation for Complex Document Processing.Proc. VLDB Endow.18, 9 (2025), 3035–3048. https: //doi.org/10.14778/3746405.3746426

    work page doi:10.14778/3746405.3746426 2025

  33. [33]

    SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

    Michael Tschannen, Alexey A. Gritsenko, Xiao Wang, Muhammad Ferjad Naeem, Ibrahim Alabdulmohsin, Nikhil Parthasarathy, Talfan Evans, Lucas Beyer, Ye Xia, Basil Mustafa, Olivier J. Hénaff, Jeremiah Harmsen, Andreas Steiner, and Xiaohua Zhai. 2025. SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense ...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2502.14786 2025

  34. [34]

    Matthias Urban and Carsten Binnig. 2024. CAESURA: Language Mod- els as Multi-Modal Query Planners. In14th Conference on Innovative Data Systems Research, CIDR 2024, Chaminade, HI, USA, January 14-17, 2024. www.cidrdb.org. https://vldb.org/cidrdb/2024/caesura-language-models-as- multi-modal-query-planners.html

    2024

  35. [35]

    Jiayi Wang and Guoliang Li. 2025. AOP: Automated and Interactive LLM Pipeline Orchestration for Answering Complex Queries. In15th Conference on Innovative Data Systems Research, CIDR 2025, Amsterdam, The Netherlands, January 19- 22, 2025. www.cidrdb.org. https://vldb.org/cidrdb/2025/aop-automated-and- Matthias Urban, Vu Huy Nguyen, Gabriele Sanmartino, Pa...

    2025

  36. [36]

    Peng Wang, Shuai Bai, Sinan Tan, Shijie Wang, Zhihao Fan, Jinze Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Yang Fan, Kai Dang, Mengfei Du, Xuancheng Ren, Rui Men, Dayiheng Liu, Chang Zhou, Jingren Zhou, and Junyang Lin. 2024. Qwen2-VL: Enhancing Vision-Language Model’s Perception of the World at Any Resolution.CoRRabs/2409.12191 (2024). https:/...

    Pith/arXiv arXiv 2024

  37. [37]

    Shihui Xu, Jiayi Wang, and Guoliang Li. 2026. Bridging the Gap: Cardinality Estimation for Semantic Queries on Unstructured Data.Proceedings of the ACM on Management of Data4, 3 (SIGMOD (2026), 1–26

    2026

  38. [38]

    Hellerstein, Sanjay Krishnan, and Ion Stoica

    Zongheng Yang, Eric Liang, Amog Kamsetty, Chenggang Wu, Yan Duan, Xi Chen, Pieter Abbeel, Joseph M. Hellerstein, Sanjay Krishnan, and Ion Stoica

  39. [39]

    VLDB Endow.13, 3 (2019), 279–292

    Deep Unsupervised Cardinality Estimation.Proc. VLDB Endow.13, 3 (2019), 279–292. https://doi.org/10.14778/3368289.3368294

    work page doi:10.14778/3368289.3368294 2019

  40. [40]

    Parameswaran

    Sepanta Zeighami, Shreya Shankar, and Aditya G. Parameswaran. 2025. Cut Costs, Not Accuracy: LLM-Powered Data Processing with Guarantees.Proc. ACM Manag. Data3, 6 (2025), 1–26. https://doi.org/10.1145/3769776

    work page doi:10.1145/3769776 2025

  41. [41]

    Xiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, and Lucas Beyer. 2023. Sig- moid Loss for Language Image Pre-Training. InIEEE/CVF International Conference on Computer Vision, ICCV 2023, Paris, France, October 1-6, 2023. IEEE, 11941–11952. https://doi.org/10.1109/ICCV51070.2023.01100

    work page doi:10.1109/iccv51070.2023.01100 2023