arxiv: 2605.10862 · v1 · submitted 2026-05-11 · 💻 cs.CL

Recognition: 2 theorem links

· Lean Theorem

RUBEN: Rule-Based Explanations for Retrieval-Augmented LLM Systems

Joel Rorseth , Parke Godfrey , Lukasz Golab , Divesh Srivastava , Jarek Szlichta

Authors on Pith no claims yet

Pith reviewed 2026-05-12 03:36 UTC · model grok-4.3

classification 💻 cs.CL

keywords rule-based explanationsretrieval-augmented LLMspruning strategiesminimal rulesLLM safetyadversarial promptsLLM interpretabilityexplanation tools

0 comments

The pith

Minimal rules found via pruning explain outputs from retrieval-augmented LLMs

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

This paper introduces RUBEN, an interactive tool that discovers a small set of rules to account for the outputs of retrieval-augmented large language models. It relies on new pruning strategies to select rules that cover every case while eliminating redundancy. The resulting rules support applications such as checking how well safety training holds up and how effective adversarial prompt injections are. A reader would care if these concise rules make the internal logic of complex LLM systems easier to inspect and control in real data applications.

Core claim

The paper claims that novel pruning strategies can efficiently identify a minimal set of rules that subsume all others, thereby explaining the outputs of retrieval-augmented LLMs. This is realized in the RUBEN tool, which allows interactive discovery of the rules and demonstrates their use in testing LLM safety, including the resiliency of safety training and the impact of adversarial prompt injections.

What carries the argument

Novel pruning strategies that identify a minimal set of rules subsuming all others to explain LLM outputs.

If this is right

The minimal rules provide complete explanations for all LLM outputs observed in the retrieval-augmented setting.
The rules enable direct testing of whether safety training in an LLM resists certain inputs.
Adversarial prompt injections can be assessed for their success or failure using the same rule set.
Interactive exploration with the tool lets users refine explanations for specific data-driven applications.

Where Pith is reading between the lines

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

If the rules remain sufficient across new inputs, the approach could reduce the effort needed to audit LLM decisions in deployed systems.
The pruning technique might apply to other generative models that combine retrieval with language generation.
Rule sets discovered this way could be combined with existing interpretability techniques to create more transparent AI pipelines.

Load-bearing premise

The pruning strategies produce rules that are both minimal and sufficient to explain all LLM outputs without omitting critical cases or introducing false explanations.

What would settle it

A dataset of LLM outputs where the pruned minimal rules leave at least one output unexplained or where reintroducing a pruned rule alters the coverage of the explanations.

Figures

Figures reproduced from arXiv: 2605.10862 by Divesh Srivastava, Jarek Szlichta, Joel Rorseth, Lukasz Golab, Parke Godfrey.

**Figure 1.** Figure 1: Using RUBEN to test adversarial prompt injections against an LLM’s safety training. RUBEN allows users to test and contrast the robustness and safety of different underlying LLMs, or versions thereof. The user configures the RAG system with a strong underlying LLM and triggers rule generation. No rules are found, indicating that the safety instructions were effective. Reconfigured with a weaker LLM, RUBEN … view at source ↗

**Figure 2.** Figure 2: The architecture of RUBEN. 22.04; the machine has an AMD Opteron 6348 Processor, 128 GB of DDR3 RAM, and GeForce RTX 2080 Ti GPU. The front-end allows users to select from three LLM systems (described in Section III). Each uses a preconfigured retriever R, LLM M, safety instructions F and predicate O. Users interact with RUBEN in two stages. In the first “retrieval” stage ( [PITH_FULL_IMAGE:figures/full_f… view at source ↗

read the original abstract

This paper demonstrates RUBEN, an interactive tool for discovering minimal rules to explain the outputs of retrieval-augmented large language models (LLMs) in data-driven applications. We leverage novel pruning strategies to efficiently identify a minimal set of rules that subsume all others. We further demonstrate novel applications of these rules for LLM safety, specifically to test the resiliency of safety training and effectiveness of adversarial prompt injections.

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.

Desk Editor's Note private letter to a colleague

RUBEN uses pruning to extract minimal rules explaining RAG outputs and applies them to safety checks, but the abstract gives little evidence on whether the rules stay complete or the tests are rigorous.

read the letter

The main takeaway is that RUBEN finds a small set of rules that cover the behavior of retrieval-augmented LLMs and then uses those rules to test safety properties such as resistance to adversarial prompts. The pruning strategies are presented as the key technical step, with the goal of producing rules that subsume all others while remaining minimal. The safety use case is a natural extension because RAG systems are already common in production and need practical ways to audit their guardrails. If the pruning works as described, it could give auditors a more manageable alternative to inspecting every retrieval or output individually. The interactive tool framing also makes sense for letting users inspect and refine the rules in real applications. The paper does a reasonable job framing the problem around data-driven RAG setups rather than generic LLMs, which keeps the scope focused. The safety demonstrations at least point to a concrete downstream task instead of stopping at explanation alone. On the soft side, the abstract stays at the level of claims without showing the pruning algorithm, coverage numbers, or any comparison to existing rule miners. It is therefore impossible to judge whether the rules actually remain sufficient or whether they drop edge cases that matter for safety. The resiliency tests would need to report specific attack success rates and false-negative rates before they can be taken as evidence of effectiveness. Without those details the contribution reads as an early prototype rather than a validated method. This paper is mainly for practitioners who already work with RAG pipelines and want rule-based auditing tools, or for researchers looking at hybrid symbolic-neural explainability. A reader who needs immediate, reproducible techniques will probably wait for more results, but the idea is clear enough that it could generate useful discussion. I would send it to peer review because the application area is timely and the core idea is simple to evaluate once the methods and experiments are laid out.

Referee Report

1 major / 1 minor

Summary. The paper introduces RUBEN, an interactive tool for discovering minimal rules to explain outputs of retrieval-augmented LLMs. It claims novel pruning strategies efficiently identify a minimal rule set that subsumes all others, and demonstrates applications of these rules to test LLM safety resiliency and adversarial prompt injection effectiveness.

Significance. If the pruning strategies provably yield minimal complete rules with rigorous coverage and safety validation, the work could advance interpretable explanations for RAG systems and offer practical tools for safety auditing. The interactive tool and safety applications represent potentially useful contributions to explainable AI, though the absence of detailed methods, metrics, or experiments in the provided text limits assessment of impact.

major comments (1)

The central claim that novel pruning strategies produce a minimal set of rules subsuming all others and explaining every RAG-LLM output lacks supporting details such as algorithm description, coverage metrics, or counter-example handling, making it impossible to verify minimality or sufficiency.

minor comments (1)

The abstract mentions 'data-driven applications' and 'interactive tool' without specifying the interface, user workflow, or example use cases.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their review of our manuscript on RUBEN. We address the major comment below and will incorporate clarifications to strengthen the presentation of our pruning strategies.

read point-by-point responses

Referee: The central claim that novel pruning strategies produce a minimal set of rules subsuming all others and explaining every RAG-LLM output lacks supporting details such as algorithm description, coverage metrics, or counter-example handling, making it impossible to verify minimality or sufficiency.

Authors: We agree that additional exposition is needed to make the claims fully verifiable from the text. The manuscript describes the pruning strategies at a high level and reports experimental outcomes, but we acknowledge the absence of pseudocode, formal coverage metrics, and explicit counter-example handling. In the revised version we will expand the methods section with the complete algorithm, definitions of minimality and completeness, quantitative coverage results across all evaluated RAG-LLM outputs, and concrete examples of how the pruning process ensures no output remains unexplained. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper describes RUBEN as an interactive tool that applies novel pruning strategies to identify minimal rule sets subsuming others for explaining RAG-LLM outputs, with applications to safety testing. No equations, fitted parameters, predictions, or derivation chains appear in the abstract or described claims. The central methodological steps are presented as new contributions without reduction to self-definition, self-citation load-bearing premises, or renaming of known results. The work is self-contained as a systems description rather than a closed mathematical loop.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no concrete free parameters, axioms, or invented entities; the description remains at the level of tool functionality and intended use cases.

pith-pipeline@v0.9.0 · 5365 in / 1055 out tokens · 43261 ms · 2026-05-12T03:36:08.153459+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We leverage novel pruning strategies to efficiently identify a minimal set of rules that subsume all others... top-down level-order traversal of the RAG source lattice... Apriori property
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

rule miner... dynamic programming approach, caching the validity of the previous level’s rules... prune inconsequential validity checks

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

57 extracted references · 57 canonical work pages · 2 internal anchors

[1]

Agrawal and R

R. Agrawal and R. Srikant, ``Fast algorithms for mining association rules in large databases,'' in PVLDB, 1994, p. 487–499

work page 1994
[2]

Cohen-Wang et al., ``Context C ite: Attributing model generation to context,'' in NeurIPS, vol

B. Cohen-Wang et al., ``Context C ite: Attributing model generation to context,'' in NeurIPS, vol. 37, 2024, pp. 95\,764--95\,807

work page 2024
[3]

Esmaelizadeh et al., ``On integrating the data-science and machine-learning pipelines for responsible AI ,'' in GUIDE-AI @ ACM SIGMOD, 2024, p

A. Esmaelizadeh et al., ``On integrating the data-science and machine-learning pipelines for responsible AI ,'' in GUIDE-AI @ ACM SIGMOD, 2024, p. 50–53

work page 2024
[4]

Rorseth et al., ``Rule-based explanations for retrieval-augmented LLM systems,'' Technical report, pp

J. Rorseth et al., ``Rule-based explanations for retrieval-augmented LLM systems,'' Technical report, pp. 1--28, 2025. [Online]. Available: https://arxiv.org/abs/2510.22689

work page arXiv 2025
[5]

5469--5472

------, `` RAGE against the machine: Retrieval-augmented LLM explanations,'' in ICDE, 2024, pp. 5469--5472

work page 2024
[6]

Rudin and Y

C. Rudin and Y. Shaposhnik, ``Globally-consistent rule-based summary-explanations for machine learning models: Application to credit-risk evaluation,'' JMLR, vol. 24, no. 16, pp. 1--44, 2023

work page 2023
[7]

Poisoning Attacks on LLMs Require a Near-constant Number of Poison Samples, October 2025

A. Souly et al., ``Poisoning attacks on LLM s require a near-constant number of poison samples,'' Technical Report, pp. 1--30, 2025. [Online]. Available: https://arxiv.org/abs/2510.07192

work page arXiv 2025
[8]

T., Singh, S

Ribeiro, Marco Tulio and Singh, Sameer and Guestrin, Carlos , title =. 2016 , isbn =. doi:10.1145/2939672.2939778 , booktitle =

work page doi:10.1145/2939672.2939778 2016
[9]

Rule-Based Explanations for Retrieval-Augmented

Joel Rorseth and Parke Godfrey and Lukasz Golab and Divesh Srivastava and Jarek Szlichta , journal=. Rule-Based Explanations for Retrieval-Augmented. 2025 , pages=

work page 2025
[10]

Vera Liao, and Rachel K

Mothilal, Ramaravind K. and Sharma, Amit and Tan, Chenhao , title =. 2020 , isbn =. doi:10.1145/3351095.3372850 , booktitle =

work page doi:10.1145/3351095.3372850 2020
[11]

Advances in Neural Information Processing Systems 30 , editor =

A Unified Approach to Interpreting Model Predictions , author =. Advances in Neural Information Processing Systems 30 , editor =. 2017 , publisher =

work page 2017
[12]

Advances in Neural Information Processing Systems , volume=

Language models are few-shot learners , author=. Advances in Neural Information Processing Systems , volume=. 2020 , publisher =

work page 2020
[13]

u tt, Kristof T. and D \

Kindermans, Pieter-Jan and Hooker, Sara and Adebayo, Julius and Alber, Maximilian and Sch \"u tt, Kristof T. and D \"a hne, Sven and Erhan, Dumitru and Kim, Been. The (Un)reliability of Saliency Methods. Explainable AI: Interpreting, Explaining and Visualizing Deep Learning. 2019. doi:10.1007/978-3-030-28954-6_14

work page doi:10.1007/978-3-030-28954-6_14 2019
[14]

Counterfactual explanations without opening the black box: Automated decisions and the GDPR , author=. Harv. JL & Tech. , volume=. 2017 , publisher=

work page 2017
[15]

JMLR , year =

Cynthia Rudin and Yaron Shaposhnik , title =. JMLR , year =

work page
[16]

PVLDB , volume=

Computing Rule-Based Explanations by Leveraging Counterfactuals , author=. PVLDB , volume=. 2022 , publisher=

work page 2022
[17]

Ribeiro, Marco Tulio and Singh, Sameer and Guestrin, Carlos , title =. Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence and Thirtieth Innovative Applications of Artificial Intelligence Conference and Eighth AAAI Symposium on Educational Advances in Artificial Intelligence , articleno =. 2018 , isbn =

work page 2018
[18]

Factual and Counterfactual Explanations for Black Box Decision Making , year=

Guidotti, Riccardo and Monreale, Anna and Giannotti, Fosca and Pedreschi, Dino and Ruggieri, Salvatore and Turini, Franco , journal=. Factual and Counterfactual Explanations for Black Box Decision Making , year=

work page
[19]

2022 , issn =

RuleXAI—A package for rule-based explanations of machine learning model , journal =. 2022 , issn =. doi:https://doi.org/10.1016/j.softx.2022.101209 , author =

work page doi:10.1016/j.softx.2022.101209 2022
[20]

Data Mining and Knowledge Discovery , volume=

Stable and actionable explanations of black-box models through factual and counterfactual rules , author=. Data Mining and Knowledge Discovery , volume=. 2024 , publisher=

work page 2024
[21]

2024 , issue_date =

Verma, Sahil and Boonsanong, Varich and Hoang, Minh and Hines, Keegan and Dickerson, John and Shah, Chirag , title =. 2024 , issue_date =. doi:10.1145/3677119 , journal =

work page doi:10.1145/3677119 2024
[22]

Cohen-Wang, Benjamin and Shah, Harshay and Georgiev, Kristian and M. Context. NeurIPS , pages =

work page
[23]

1994 , booktitle =

Agrawal, Rakesh and Srikant, Ramakrishnan , title =. 1994 , booktitle =

work page 1994
[24]

2024 , volume=

Rorseth, Joel and Godfrey, Parke and Golab, Lukasz and Srivastava, Divesh and Szlichta, Jaroslaw , booktitle=. 2024 , volume=

work page 2024
[25]

CREDENCE: Counterfactual Explanations for Document Ranking , year=

Rorseth, Joel and Godfrey, Parke and Golab, Lukasz and Kargar, Mehdi and Srivastava, Divesh and Szlichta, Jaroslaw , booktitle=. CREDENCE: Counterfactual Explanations for Document Ranking , year=

work page
[26]

2024 , isbn =

Xu, Zhichao and Lamba, Hemank and Ai, Qingyao and Tetreault, Joel and Jaimes, Alex , title =. 2024 , isbn =. doi:10.1145/3664190.3672508 , booktitle =

work page doi:10.1145/3664190.3672508 2024
[27]

, booktitle =

Yang, Zhilin and Qi, Peng and Zhang, Saizheng and Bengio, Yoshua and Cohen, William and Salakhutdinov, Ruslan and Manning, Christopher D. H otpot QA : A Dataset for Diverse, Explainable Multi-hop Question Answering. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. 2018. doi:10.18653/v1/D18-1259

work page doi:10.18653/v1/d18-1259 2018
[28]

NPJ digital medicine , volume=

Medical large language models are susceptible to targeted misinformation attacks , author=. NPJ digital medicine , volume=. 2024 , publisher=

work page 2024
[29]

First Conference on Language Modeling , year=

Source-Aware Training Enables Knowledge Attribution in Language Models , author=. First Conference on Language Modeling , year=

work page
[30]

Towards Automated Circuit Discovery for Mechanistic Interpretability , volume =

Conmy, Arthur and Mavor-Parker, Augustine and Lynch, Aengus and Heimersheim, Stefan and Garriga-Alonso, Adri\`. Towards Automated Circuit Discovery for Mechanistic Interpretability , volume =. Advances in Neural Information Processing Systems , editor =. 2023 , publisher =

work page 2023
[31]

Can Large Language Models Be an Alternative to Human Evaluations?

Chiang, Cheng-Han and Lee, Hung-yi. Can Large Language Models Be an Alternative to Human Evaluations?. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023. doi:10.18653/v1/2023.acl-long.870

work page doi:10.18653/v1/2023.acl-long.870 2023
[32]

and Zhang, Hao and Gonzalez, Joseph E

Zheng, Lianmin and Chiang, Wei-Lin and Sheng, Ying and Zhuang, Siyuan and Wu, Zhanghao and Zhuang, Yonghao and Lin, Zi and Li, Zhuohan and Li, Dacheng and Xing, Eric P. and Zhang, Hao and Gonzalez, Joseph E. and Stoica, Ion , title =. 2023 , booktitle =

work page 2023
[33]

1995 , publisher=

Foundations of databases , author=. 1995 , publisher=

work page 1995
[34]

International Journal on Digital Libraries , volume=

Citation recommendation: approaches and datasets , author=. International Journal on Digital Libraries , volume=. 2020 , publisher=

work page 2020
[35]

CiteME: Can Language Models Accurately Cite Scientific Claims? , volume =

Press, Ori and Hochlehnert, Andreas and Prabhu, Ameya and Udandarao, Vishaal and Press, Ofir and Bethge, Matthias , booktitle =. CiteME: Can Language Models Accurately Cite Scientific Claims? , volume =

work page
[36]

Verifiable Generation with Subsentence-Level Fine-Grained Citations

Cao, Shuyang and Wang, Lu. Verifiable Generation with Subsentence-Level Fine-Grained Citations. Findings of the Association for Computational Linguistics: ACL 2024. 2024. doi:10.18653/v1/2024.findings-acl.920

work page doi:10.18653/v1/2024.findings-acl.920 2024
[37]

Do Language Models Know When They ' re Hallucinating References?

Agrawal, Ayush and Suzgun, Mirac and Mackey, Lester and Kalai, Adam. Do Language Models Know When They ' re Hallucinating References?. Findings of the Association for Computational Linguistics: EACL 2024. 2024

work page 2024
[38]

This Reference Does Not Exist: An Exploration of LLM Citation Accuracy and Relevance

Byun, Courtni and Vasicek, Piper and Seppi, Kevin. This Reference Does Not Exist: An Exploration of LLM Citation Accuracy and Relevance. Proceedings of the Third Workshop on Bridging Human--Computer Interaction and Natural Language Processing. 2024. doi:10.18653/v1/2024.hcinlp-1.3

work page doi:10.18653/v1/2024.hcinlp-1.3 2024
[39]

The Professional Geographer , volume =

Terence Day , title =. The Professional Geographer , volume =. 2023 , publisher =

work page 2023
[40]

Galactica: A Large Language Model for Science

Galactica: A large language model for science , author=. arXiv preprint arXiv:2211.09085 , year=

work page internal anchor Pith review arXiv
[41]

Enabling Large Language Models to Generate Text with Citations

Gao, Tianyu and Yen, Howard and Yu, Jiatong and Chen, Danqi. Enabling Large Language Models to Generate Text with Citations. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023. doi:10.18653/v1/2023.emnlp-main.398

work page doi:10.18653/v1/2023.emnlp-main.398 2023
[42]

Learning to Generate Answers with Citations via Factual Consistency Models

Aly, Rami and Tang, Zhiqiang and Tan, Samson and Karypis, George. Learning to Generate Answers with Citations via Factual Consistency Models. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024. doi:10.18653/v1/2024.acl-long.641

work page doi:10.18653/v1/2024.acl-long.641 2024
[43]

C ite F ix: Enhancing RAG Accuracy Through Post-Processing Citation Correction

Maheshwari, Harsh and Tenneti, Srikanth and Nakkiran, Alwarappan. C ite F ix: Enhancing RAG Accuracy Through Post-Processing Citation Correction. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track). 2025. doi:10.18653/v1/2025.acl-industry.23

work page doi:10.18653/v1/2025.acl-industry.23 2025
[44]

HLM-Cite: Hybrid Language Model Workflow for Text-based Scientific Citation Prediction , volume =

Hao, Qianyue and Fan, Jingyang and Xu, Fengli and Yuan, Jian and Li, Yong , booktitle =. HLM-Cite: Hybrid Language Model Workflow for Text-based Scientific Citation Prediction , volume =

work page
[45]

WebGPT: Browser-assisted question-answering with human feedback

Webgpt: Browser-assisted question-answering with human feedback , author=. arXiv preprint arXiv:2112.09332 , year=

work page internal anchor Pith review Pith/arXiv arXiv
[46]

Citation-Enhanced Generation for LLM -based Chatbots

Li, Weitao and Li, Junkai and Ma, Weizhi and Liu, Yang. Citation-Enhanced Generation for LLM -based Chatbots. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024. doi:10.18653/v1/2024.acl-long.79

work page doi:10.18653/v1/2024.acl-long.79 2024
[47]

InterpBench: Semi-Synthetic Transformers for Evaluating Mechanistic Interpretability Techniques , volume =

Gupta, Rohan and Arcuschin, Iv\'. InterpBench: Semi-Synthetic Transformers for Evaluating Mechanistic Interpretability Techniques , volume =. Advances in Neural Information Processing Systems , editor =

work page
[48]

Grokking of Implicit Reasoning in Transformers: A Mechanistic Journey to the Edge of Generalization , volume =

Wang, Boshi and Yue, Xiang and Su, Yu and Sun, Huan , booktitle =. Grokking of Implicit Reasoning in Transformers: A Mechanistic Journey to the Edge of Generalization , volume =

work page
[49]

Slack, S

Slack, Dylan and Hilgard, Sophie and Jia, Emily and Singh, Sameer and Lakkaraju, Himabindu , title =. 2020 , isbn =. doi:10.1145/3375627.3375830 , booktitle =

work page doi:10.1145/3375627.3375830 2020
[50]

Problems With SHAP and LIME in Interpretable AI for Education: A Comparative Study of Post-Hoc Explanations and Neural-Symbolic Rule Extraction , year=

Hooshyar, Danial and Yang, Yeongwook , journal=. Problems With SHAP and LIME in Interpretable AI for Education: A Comparative Study of Post-Hoc Explanations and Neural-Symbolic Rule Extraction , year=

work page
[51]

Decisions, Counterfactual Explanations and Strategic Behavior , volume =

Tsirtsis, Stratis and Gomez Rodriguez, Manuel , booktitle =. Decisions, Counterfactual Explanations and Strategic Behavior , volume =

work page
[52]

and Agrawal, Rakesh , title =

Bayardo, Roberto J. and Agrawal, Rakesh , title =. Proceedings of the Fifth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages =. 1999 , isbn =. doi:10.1145/312129.312219 , publisher =

work page doi:10.1145/312129.312219 1999
[53]

Mining association rules between sets of items in large databases , year =

Agrawal, Rakesh and Imieli\'. Mining association rules between sets of items in large databases , year =. Proceedings of the 1993 ACM SIGMOD International Conference on Management of Data , pages =. doi:10.1145/170035.170072 , publisher =

work page doi:10.1145/170035.170072 1993
[54]

Mining association rules between sets of items in large databases , year =

Agrawal, Rakesh and Imieli\'. Mining association rules between sets of items in large databases , year =. doi:10.1145/170036.170072 , journal =

work page doi:10.1145/170036.170072
[55]

Bayan , title =

Hameed, Isha and Sharpe, Samuel and Barcklow, Daniel and Au-Yeung, Justin and Verma, Sahil and Huang, Jocelyn and Barr, Brian and Bruss, C. Bayan , title =. 2022 , maintitle =

work page 2022
[56]

Poisoning Attacks on

Souly, Alexandra and Rando, Javier and Chapman, Ed and Davies, Xander and Hasircioglu, Burak and Shereen, Ezzeldin and Mougan, Carlos and Mavroudis, Vasilios and Jones, Erik and Hicks, Chris and Nicholas Carlini and Yarin Gal and Robert Kirk , journal=. Poisoning Attacks on. 2025 , pages=

work page 2025
[57]

GUIDE-AI @ ACM SIGMOD , pages =

Esmaelizadeh, Armin and Rorseth, Joel and Yu, Andy and Godfrey, Parke and Golab, Lukasz and Srivastava, Divesh and Szlichta, Jaroslaw and Taghva, Kazem , title =. GUIDE-AI @ ACM SIGMOD , pages =. 2024 , isbn =

work page 2024