pith. sign in

arxiv: 2604.19251 · v1 · submitted 2026-04-21 · 💻 cs.LO · cs.AI

Streamliners for Answer Set Programming

Florentina Voboril , Martin Gebser , Stefan Szeider , Alice Tarzariol This is my paper

Pith reviewed 2026-05-10 02:01 UTC · model grok-4.3

classification 💻 cs.LO cs.AI
keywords answer set programmingstreamliner constraintslarge language modelsconstraint generationsolver performancevirtual best encodingASP benchmarks
0
0 comments X

The pith

Large language models can generate streamliner constraints that speed up Answer Set Programming solvers by up to five times on competition benchmarks.

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

The paper adapts an LLM-based method for creating streamliner constraints to Answer Set Programming. Given an existing ASP encoding and a handful of small training instances, multiple LLMs are prompted to suggest additional constraints that rule out unproductive parts of the search space. Candidates are filtered for syntactic validity, preservation of satisfiability, and performance gains on the training set. A virtual best encoding then selects, for each new instance, the fastest version among the original encoding and the surviving streamlined variants. On three standard benchmarks the selection process delivers consistent speedups while revealing that different models contribute distinct semantic ideas rather than simple rephrasings.

Core claim

Given an ASP encoding and a few small training instances, multiple LLMs are prompted to propose candidate streamliner constraints. These candidates are filtered to remove those that cause syntax errors, make satisfiable instances unsatisfiable, or slow down all training instances. The remaining streamliners are combined with the original encoding, and a virtual best encoding chooses the fastest one for each test instance. On the Partner Units Problem, Sokoban, and Towers of Hanoi benchmarks, this approach achieves speedups of up to 4-5 times compared to the original encoding alone, with different LLMs producing semantically diverse constraints.

What carries the argument

The virtual best encoding that, for each instance, selects the fastest among the original ASP encoding and its LLM-generated streamlined variants after filtering for syntax, satisfiability, and training performance.

If this is right

  • Automatic generation of streamliners becomes feasible for ASP problems without requiring manual expert crafting of constraints.
  • Using multiple distinct LLMs increases the variety of proposed constraints and the likelihood that at least one effective variant survives filtering.
  • The virtual best encoding can be applied instance-by-instance during solving to exploit the strengths of different streamlined variants.
  • The observed semantic diversity among LLM proposals indicates that the method extracts genuine structural features rather than surface-level syntactic changes.

Where Pith is reading between the lines

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

  • If the method scales, it could lower the barrier for non-experts to obtain high-performance ASP encodings for new combinatorial problems.
  • Iterative prompting on intermediate solving results might allow dynamic refinement of streamliners during a single run.
  • The filtering pipeline could be extended with machine-learning predictors of generalization to reduce reliance on full training-set evaluation.
  • Similar LLM-driven streamlining might transfer to other declarative paradigms such as SAT or CP once the filtering criteria are adapted.

Load-bearing premise

Candidate streamliners generated from a few small training instances will generalize to larger unseen instances without introducing errors or performance regressions on the target benchmark set.

What would settle it

Applying the filtered streamliners to a fresh collection of large instances from the same three benchmarks and measuring whether any render satisfiable problems unsatisfiable or whether the virtual best encoding fails to improve on the original solver time.

Figures

Figures reproduced from arXiv: 2604.19251 by Alice Tarzariol, Florentina Voboril, Martin Gebser, Stefan Szeider.

Figure 1
Figure 1. Figure 1: Fully automated pipeline of the approach [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Used prompt • The prompt for the original StreamLLM approach tells the LLM how to deal with feedback on previously provided constraints. Since their experiments showed that this adaptive variant is not working significantly better than a variant without feedback, we omitted this part. Typically, the LLM returns single constraints. However, it is also possible that it returns multiple constraints or new rul… view at source ↗
Figure 3
Figure 3. Figure 3: Results for PUP for run 1 (left) and run 2 (right) [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Results for Sokoban for run 1 (left) and run 2 (right) [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Results for Towers of Hanoi for run 1 (left) and run 2 (right) [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
read the original abstract

Streamliner constraints reduce the search space of combinatorial problems by ruling out portions of the solution space. We adapt the StreamLLM approach, which uses Large Language Models (LLMs) to generate streamliners for Constraint Programming, to Answer Set Programming (ASP). Given an ASP encoding and a few small training instances, we prompt multiple LLMs to propose candidate constraints. Candidates that cause syntax errors, render satisfiable instances unsatisfiable, or degrade performance on all training instances are discarded. The surviving streamliners are evaluated together with the original encoding, and we report results for a virtual best encoding (VBE) that, for each instance, selects the fastest among the original encoding and its streamlined variants. On three ASP Competition benchmarks (Partner Units Problem, Sokoban, Towers of Hanoi), the VBE achieves speedups of up to 4--5x over the original encoding. Different LLMs produce semantically diverse constraints, not mere syntactic variations, indicating that the approach captures genuine problem structure.

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 / 2 minor

Summary. The paper adapts the StreamLLM approach to Answer Set Programming by prompting multiple LLMs to generate candidate streamliner constraints from a small set of training instances. Candidates are filtered to remove those causing syntax errors, rendering satisfiable training instances unsatisfiable, or degrading performance on all training instances. Surviving streamliners are combined with the original encoding, and results are reported for a virtual best encoding (VBE) that selects the fastest variant per instance. On three ASP Competition benchmarks (Partner Units Problem, Sokoban, Towers of Hanoi), the VBE yields speedups of up to 4-5x, with different LLMs producing semantically diverse rather than merely syntactic constraints.

Significance. If the central performance claims hold after verification, the work provides a practical demonstration that LLMs can generate useful, diverse streamliners for ASP encodings, offering measurable speedups on standard combinatorial benchmarks. The VBE methodology and use of independent competition instances make the results falsifiable and reproducible in principle, which could encourage further hybrid LLM-solver techniques in logic programming and constraint solving.

major comments (2)
  1. [Abstract] Abstract: The filtering protocol discards candidates that render satisfiable training instances unsatisfiable, yet the reported 4-5x speedups are measured on the full ASP Competition benchmark instances, which are larger than the training set. No evidence is given that retained streamliners were re-checked for soundness (preservation of satisfiability) or absence of performance regressions on these test instances. This verification step is load-bearing for the validity of the VBE speedup claims, as an unsound streamliner selected by the VBE oracle would invalidate the measured improvements.
  2. [Abstract] Abstract / Evaluation: The abstract states concrete speedups but provides no details on training-instance sizes relative to test instances, the number of instances per benchmark, the number of independent runs, or any statistical tests supporting the speedup figures. These omissions make it impossible to assess whether the 4-5x gains are robust or could be artifacts of instance selection or variance.
minor comments (2)
  1. The abstract would benefit from naming the specific LLMs used and briefly outlining the prompt structure to aid reproducibility.
  2. Notation for the VBE construction and the filtering criteria could be made more explicit (e.g., via pseudocode or a small table) rather than described only in prose.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thoughtful and detailed report. The comments identify important gaps in the presentation of soundness guarantees and experimental details. We respond to each major comment below and indicate the revisions we will make.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The filtering protocol discards candidates that render satisfiable training instances unsatisfiable, yet the reported 4-5x speedups are measured on the full ASP Competition benchmark instances, which are larger than the training set. No evidence is given that retained streamliners were re-checked for soundness (preservation of satisfiability) or absence of performance regressions on these test instances. This verification step is load-bearing for the validity of the VBE speedup claims, as an unsound streamliner selected by the VBE oracle would invalidate the measured improvements.

    Authors: We agree that the absence of explicit soundness re-verification on the competition test instances is a limitation of the current presentation. Soundness and performance filtering were applied only to the small training instances, as described in the manuscript. The virtual best encoding always retains the original encoding as a fallback option, so any reported VBE solution is produced by a variant that was at least sound on training data. Nevertheless, we acknowledge that an unsound streamliner could in principle be selected for a test instance. In the revised manuscript we will add a dedicated paragraph in the evaluation section that (a) explicitly states the scope of the soundness checks, (b) discusses the risk for test instances, and (c) reports a post-hoc manual inspection of the selected streamliners on a sample of competition instances to confirm that no satisfiable instance was rendered unsatisfiable. We will also qualify the abstract claims accordingly. revision: yes

  2. Referee: [Abstract] Abstract / Evaluation: The abstract states concrete speedups but provides no details on training-instance sizes relative to test instances, the number of instances per benchmark, the number of independent runs, or any statistical tests supporting the speedup figures. These omissions make it impossible to assess whether the 4-5x gains are robust or could be artifacts of instance selection or variance.

    Authors: We accept that the abstract is insufficiently self-contained on these points. The full manuscript contains the relevant numbers in the experimental setup and results sections, but they are not summarized in the abstract. In the revision we will expand the abstract to state: (i) the sizes of the training instances relative to the competition instances, (ii) the exact number of instances per benchmark drawn from the ASP Competition, (iii) that results are reported from single deterministic runs (standard practice for VBE comparisons), and (iv) that the observed speedups are consistent across all instances of each benchmark. We will also add a short note on the lack of formal statistical testing, explaining that the VBE selection is deterministic once the streamliners are fixed. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical filtering on training instances with independent benchmark evaluation

full rationale

The paper presents an empirical pipeline: LLMs generate candidate streamliners from a few small training instances; invalid or non-improving candidates are filtered using syntax checks, satisfiability preservation, and runtime on those same training instances; surviving variants are then combined into a virtual best encoding whose speedups are measured directly on three separate ASP Competition benchmark collections. No equations, self-definitional relations, fitted parameters renamed as predictions, or load-bearing self-citations appear in the derivation. The reported speedups are therefore independent measurements on held-out data rather than reductions to the training inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on standard ASP semantics and existing LLM capabilities; no new free parameters, axioms, or invented entities are introduced beyond domain-standard assumptions.

axioms (1)
  • standard math Standard ASP semantics and solver behavior hold for the encodings used.
    Invoked implicitly when claiming that generated constraints preserve satisfiability.

pith-pipeline@v0.9.0 · 5473 in / 1194 out tokens · 39087 ms · 2026-05-10T02:01:23.145836+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

43 extracted references · 40 canonical work pages

  1. [1]

    In:Integration of AI and OR Techniques in Constraint Programming, CPAIOR 2011,Lecture Notes in Computer Science6697, Springer, pp

    Markus Aschinger, Conrad Drescher, Gerhard Friedrich, Georg Gottlob, Peter Jeavons, Anna Ryabokon & Evgenij Thorstensen (2011):Optimization Methods for the Partner Units Problem. In:Integration of AI and OR Techniques in Constraint Programming, CPAIOR 2011,Lecture Notes in Computer Science6697, Springer, pp. 4–19, doi:10.1007/978-3-642-21311-3_4

    work page doi:10.1007/978-3-642-21311-3_4 2011

  2. [2]

    Bomanson, M

    J. Bomanson, M. Gebser, T. Janhunen, B. Kaufmann & T. Schaub (2016):Answer Set Programming Modulo Acyclicity.Fundamenta Informaticae147(1), pp. 63–91, doi:10.3233/FI-2016-1398

    work page doi:10.3233/fi-2016-1398 2016

  3. [3]

    294–309, doi:10.1017/S1471068419000450

    Francesco Calimeri, Wolfgang Faber, Martin Gebser, Giovambattista Ianni, Roland Kaminski, Thomas Krennwallner, Nicola Leone, Marco Maratea, Francesco Ricca & Torsten Schaub (2020):ASP- Core-2 Input Language Format.Theory and Practice of Logic Programming20(2), pp. 294–309, doi:10.1017/S1471068419000450

    work page doi:10.1017/s1471068419000450 2020

  4. [4]

    117–135, doi:10.1017/S1471068412000105

    Francesco Calimeri, Giovambattista Ianni & Francesco Ricca (2014):The third open answer set programming competition.Theory and Practice of Logic Programming14(1), pp. 117–135, doi:10.1017/S1471068412000105

    work page doi:10.1017/s1471068412000105 2014

  5. [5]

    848–864, doi:10.1017/s147106842300025x

    Paola Cappanera, Marco Gavanelli, Maddalena Nonato & Marco Roma (2023):Logic-Based Benders De- composition in Answer Set Programming for Chronic Outpatients Scheduling.Theory and Practice of Logic Programming23(4), pp. 848–864, doi:10.1017/s147106842300025x

    work page doi:10.1017/s147106842300025x 2023

  6. [6]

    In:Proceedings of the 21st International Conference on Principles of Knowledge Representation and Reasoning, KR 2024, doi:10.24963/kr.2024/78

    Erica Coppolillo, Francesco Calimeri, Giuseppe Manco, Simona Perri & Francesco Ricca (2024):LLASP: Fine-tuning Large Language Models for Answer Set Programming. In:Proceedings of the 21st International Conference on Principles of Knowledge Representation and Reasoning, KR 2024, doi:10.24963/kr.2024/78

    work page doi:10.24963/kr.2024/78 2024

  7. [7]

    CoRRabs/1608.08447

    Jo Devriendt & Bart Bogaerts (2016):BreakID: Static Symmetry Breaking for ASP (System Description). CoRRabs/1608.08447. arXiv:1608.08447

    work page arXiv 2016

  8. [8]

    636–652, doi:10.1017/S1471068416000508

    Jo Devriendt, Bart Bogaerts, Maurice Bruynooghe & Marc Denecker (2016):On local domain sym- metry for model expansion.Theory and Practice of Logic Programming16(5-6), pp. 636–652, doi:10.1017/S1471068416000508

    work page doi:10.1017/s1471068416000508 2016

  9. [9]

    24(2), pp

    Christian Drescher, Oana Tifrea & Toby Walsh (2011):Symmetry-breaking answer set solving.AI Commun. 24(2), pp. 177–194, doi:10.3233/AIC-2011-0495

    work page doi:10.3233/aic-2011-0495 2011

  10. [10]

    Gebser, A

    M. Gebser, A. Harrison, R. Kaminski, V . Lifschitz & T. Schaub (2015):Abstract Gringo.Theory and Practice of Logic Programming15(4-5), pp. 449–463, doi:10.1017/S1471068415000150

    work page doi:10.1017/s1471068415000150 2015

  11. [11]

    Gebser, R

    M. Gebser, R. Kaminski, B. Kaufmann & T. Schaub (2012):Answer Set Solving in Practice. Morgan and Claypool Publishers, doi:10.1007/978-3-031-01561-8

    work page doi:10.1007/978-3-031-01561-8 2012

  12. [12]

    In:Technical Communications of the 32nd International Conference on Logic Programming, ICLP 2016,OASIcs52, Schloss Dagstuhl - Leibniz- Zentrum für Informatik, pp

    Martin Gebser, Roland Kaminski, Benjamin Kaufmann, Patrick Lühne, Javier Romero & Torsten Schaub (2016):Answer Set Solving with Generalized Learned Constraints. In:Technical Communications of the 32nd International Conference on Logic Programming, ICLP 2016,OASIcs52, Schloss Dagstuhl - Leibniz- Zentrum für Informatik, pp. 9:1–9:15, doi:10.4230/OASIcs.ICLP.2016.9

    work page doi:10.4230/oasics.iclp.2016.9 2016

  13. [13]

    27–82, doi:10.1017/S1471068418000054

    Martin Gebser, Roland Kaminski, Benjamin Kaufmann & Torsten Schaub (2019):Multi-shot ASP solving with clingo.Theory and Practice of Logic Programming19(1), pp. 27–82, doi:10.1017/S1471068418000054

    work page doi:10.1017/s1471068418000054 2019

  14. [14]

    Intell.187, pp

    Martin Gebser, Benjamin Kaufmann & Torsten Schaub (2012):Conflict-driven answer set solving: From theory to practice.Artif. Intell.187, pp. 52–89, doi:10.1016/j.artint.2012.04.001

    work page doi:10.1016/j.artint.2012.04.001 2012

  15. [15]

    Comput.9(3/4), pp

    Michael Gelfond & Vladimir Lifschitz (1991):Classical Negation in Logic Programs and Disjunctive Databases.New Gener. Comput.9(3/4), pp. 365–386, doi:10.1007/BF03037169. 14Streamliners for Answer Set Programming

    work page doi:10.1007/bf03037169 1991

  16. [16]

    Gomes & Meinolf Sellmann (2004):Streamlined Constraint Reasoning

    Carla P. Gomes & Meinolf Sellmann (2004):Streamlined Constraint Reasoning. In Mark Wallace, editor: Principles and Practice of Constraint Programming - CP 2004,Lecture Notes in Computer Science3258, Springer, pp. 274–289, doi:10.1007/978-3-540-30201-8_22

    work page doi:10.1007/978-3-540-30201-8_22 2004

  17. [17]

    Hoos, Roland Kaminski, Marius Lindauer & Torsten Schaub (2015):aspeed: Solver schedul- ing via answer set programming.Theory and Practice of Logic Programming15(1), pp

    Holger H. Hoos, Roland Kaminski, Marius Lindauer & Torsten Schaub (2015):aspeed: Solver schedul- ing via answer set programming.Theory and Practice of Logic Programming15(1), pp. 117–142, doi:10.1017/S1471068414000015

    work page doi:10.1017/s1471068414000015 2015

  18. [18]

    In:21st International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2022, IFAAMAS, pp

    Matej Husár, Jirí Svancara, Philipp Obermeier, Roman Barták & Torsten Schaub (2022):Reduction- based Solving of Multi-agent Pathfinding on Large Maps Using Graph Pruning. In:21st International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2022, IFAAMAS, pp. 624–632, doi:10.5555/3535850.3535921

    work page doi:10.5555/3535850.3535921 2022

  19. [19]

    2023 , month =

    Adam Ishay, Zhun Yang & Joohyung Lee (2023):Leveraging Large Language Models to Generate Answer Set Programs. In:Proceedings of the 20th International Conference on Principles of Knowledge Represen- tation and Reasoning, KR 2023, pp. 374–383, doi:10.24963/kr.2023/37

    work page doi:10.24963/kr.2023/37 2023

  20. [20]

    387–414, doi:10.1017/S1471068422000011

    Mark Law (2023):Conflict-Driven Inductive Logic Programming.Theory and Practice of Logic Program- ming23(2), pp. 387–414, doi:10.1017/S1471068422000011

    work page doi:10.1017/s1471068422000011 2023

  21. [21]

    In:Proceedings of the AAAI Conference on Artificial Intelligence, AAAI 2024, AAAI Press, pp

    Kevin Leo, Graeme Gange, Maria Garcia de la Banda & Mark Wallace (2024):Automatic Core-Guided Reformulation via Constraint Explanation and Condition Learning. In:Proceedings of the AAAI Conference on Artificial Intelligence, AAAI 2024, AAAI Press, pp. 8065–8072, doi:10.1609/aaai.v38i8.28645

    work page doi:10.1609/aaai.v38i8.28645 2024

  22. [22]

    In: Computer Vision – ECCV 2018

    Vladimir Lifschitz, Patrick Lühne & Torsten Schaub (2019):Verifying Strong Equivalence of Programs in the Input Language of gringo. In:Logic Programming and Nonmonotonic Reasoning - 15th International Con- ference, LPNMR 2019,Lecture Notes in Computer Science11481, Springer, pp. 270–283, doi:10.1007/978- 3-030-20528-7_20

    work page doi:10.1007/978- 2019

  23. [23]

    841–868, doi:10.1017/S1471068413000094

    Marco Maratea, Luca Pulina & Francesco Ricca (2014):A Multi-Engine Approach to Answer-Set Program- ming.Theory and Practice of Logic Programming14(6), pp. 841–868, doi:10.1017/S1471068413000094

    work page doi:10.1017/s1471068413000094 2014

  24. [24]

    197–224, doi:10.1017/s1471068424000462

    Javier Romero, Torsten Schaub & Klaus Strauch (2025):On the Generalization of Learned Constraints for ASP Solving in Temporal Domains.Theory and Practice of Logic Programming25(2), pp. 197–224, doi:10.1017/s1471068424000462

    work page doi:10.1017/s1471068424000462 2025

  25. [25]

    Sakallah (2009):Symmetry and Satisfiability

    Karem A. Sakallah (2009):Symmetry and Satisfiability. In Armin Biere, Marijn Heule, Hans van Maaren & Toby Walsh, editors:Handbook of Satisfiability,Frontiers in Artificial Intelligence and Applications185, IOS Press, pp. 289–338, doi:10.3233/978-1-58603-929-5-289

    work page doi:10.3233/978-1-58603-929-5-289 2009

  26. [26]

    Borroto Santana, Irfan Kareem & Francesco Ricca (2024):Towards Automatic Composition of ASP Programs from Natural Language Specifications

    Manuel A. Borroto Santana, Irfan Kareem & Francesco Ricca (2024):Towards Automatic Composition of ASP Programs from Natural Language Specifications. In:Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, IJCAI 2024, ijcai.org, pp. 6198–6206. Available athttps: //www.ijcai.org/proceedings/2024/685

    2024

  27. [27]

    arXiv:2412.14814

    Arseny Skryagin, Daniel Ochs, Phillip Deibert, Simon Kohaut, Devendra Singh Dhami & Kristian Ker- sting (2024):Answer Set Networks: Casting Answer Set Programming into Deep Learning.CoRR abs/2412.14814, doi:10.48550/arXiv.2412.14814. arXiv:2412.14814

    work page doi:10.48550/arxiv.2412.14814 2024

  28. [28]

    Intell.319, p

    Patrick Spracklen, Nguyen Dang, Özgür Akgün & Ian Miguel (2023):Automated streamliner portfolios for constraint satisfaction problems.Artif. Intell.319, p. 103915, doi:10.1016/j.artint.2023.103915

    work page doi:10.1016/j.artint.2023.103915 2023

  29. [29]

    Available athttps://doi.org/10.5281/zenodo.18062939

    Stefan Szeider (2025):ASP-Bench: Problems, Ground Truths, and Solutions, doi:10.5281/zenodo.18062939. Available athttps://doi.org/10.5281/zenodo.18062939

    work page doi:10.5281/zenodo.18062939 2025

  30. [30]

    In: 28th International Conference on Theory and Applications of Satisfiability Testing, SAT 2025,LIPIcs341, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, pp

    Stefan Szeider (2025):Bridging Language Models and Symbolic Solvers via the Model Context Protocol. In: 28th International Conference on Theory and Applications of Satisfiability Testing, SAT 2025,LIPIcs341, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, pp. 30:1–30:12, doi:10.4230/LIPIcs.SAT.2025.30

    work page doi:10.4230/lipics.sat.2025.30 2025

  31. [31]

    Learn.111(4), pp

    Alice Tarzariol, Martin Gebser & Konstantin Schekotihin (2022):Lifting symmetry breaking constraints with inductive logic programming.Mach. Learn.111(4), pp. 1303–1326, doi:10.1007/s10994-022-06146-3

    work page doi:10.1007/s10994-022-06146-3 2022

  32. [32]

    In Brian Williams, Yiling Chen & Jennifer Neville, F

    Alice Tarzariol, Martin Gebser, Konstantin Schekotihin & Mark Law (2023):Learning to Break Symmetries for Efficient Optimization in Answer Set Programming. In Brian Williams, Yiling Chen & Jennifer Neville, F. V oboril, M. Gebser, S. Szeider & A. Tarzariol15 editors:Thirty-Seventh AAAI Conference on Artificial Intelligence, AAAI 2023, AAAI Press, pp. 6541...

    work page doi:10.1609/aaai.v37i5.25804 2023

  33. [33]

    606–622, doi:10.1017/S1471068422000151

    Alice Tarzariol, Konstantin Schekotihin, Martin Gebser & Mark Law (2022):Efficient Lifting of Symmetry Breaking Constraints for Complex Combinatorial Problems.Theory and Practice of Logic Programming 22(4), pp. 606–622, doi:10.1017/S1471068422000151

    work page doi:10.1017/s1471068422000151 2022

  34. [34]

    799– 814, doi:10.1017/S1471068420000368

    Richard Taupe, Antonius Weinzierl & Gerhard Friedrich (2020):Conflict Generalisation in ASP: Learning Correct and Effective Non-Ground Constraints.Theory and Practice of Logic Programming20(5), pp. 799– 814, doi:10.1017/S1471068420000368

    work page doi:10.1017/s1471068420000368 2020

  35. [35]

    Erich Christian Teppan, Gerhard Friedrich & Georg Gottlob (2016):Tractability frontiers of the partner units configuration problem.J. Comput. Syst. Sci.82(5), pp. 739–755, doi:10.1016/j.jcss.2015.12.004

    work page doi:10.1016/j.jcss.2015.12.004 2016

  36. [36]

    Florentina V oboril, Martin Gebser, Stefan Szeider & Alice Tarzariol (2026):Code and Instances for the Paper: Streamliners for Answer Set Programming, doi:10.5281/zenodo.18378760

    work page doi:10.5281/zenodo.18378760 2026

  37. [37]

    Florentina V oboril, Vaidyanathan Peruvemba Ramaswamy & Stefan Szeider (2025):Balancing Latin Rectan- gles with LLM-Generated Streamliners. In Maria Garcia de la Banda, editor:31st International Conference on Principles and Practice of Constraint Programming, CP 2025,LIPIcs340, Schloss Dagstuhl - Leibniz- Zentrum für Informatik, pp. 36:1–36:17, doi:10.423...

    work page doi:10.4230/lipics.cp.2025.36 2025

  38. [38]

    Florentina V oboril, Vaidyanathan Peruvemba Ramaswamy & Stefan Szeider (2025):Generating Streamlin- ing Constraints with Large Language Models.J. Artif. Intell. Res.84, doi:10.1613/jair.1.18965

    work page doi:10.1613/jair.1.18965 2025

  39. [39]

    In:Proceedings of the Twenty-Sixth AAAI Conference on Artificial Intelligence, AAAI 2012, AAAI Press, doi:10.1609/AAAI.V26I1.8437

    Toby Walsh (2012):Symmetry Breaking Constraints: Recent Results. In:Proceedings of the Twenty-Sixth AAAI Conference on Artificial Intelligence, AAAI 2012, AAAI Press, doi:10.1609/AAAI.V26I1.8437

    work page doi:10.1609/aaai.v26i1.8437 2012

  40. [40]

    In: Advances in Neural Information Processing Systems, 35, Curran Associates, Inc., pp

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Richter, Fei Xia, Ed Chi, Quoc V Le & Denny Zhou (2022):Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. In: Advances in Neural Information Processing Systems, 35, Curran Associates, Inc., pp. 24824–24837

    2022

  41. [41]

    In:2nd Workshop on Ground- ing and Transformations for Theories With Variables (GTTV’13), p

    Antonius Weinzierl (2013):Learning non-ground rules for answer-set solving. In:2nd Workshop on Ground- ing and Transformations for Theories With Variables (GTTV’13), p. 13

    2013

  42. [42]

    NeurASP: Embracing Neural Networks into Answer Set Programming , booktitle =

    Zhun Yang, Adam Ishay & Joohyung Lee (2020):NeurASP: Embracing Neural Networks into Answer Set Programming. In:Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, IJCAI 2020, ijcai.org, pp. 1755–1762, doi:10.24963/ijcai.2020/243

    work page doi:10.24963/ijcai.2020/243 2020

  43. [43]

    In:PADL 2023, Lecture Notes in Computer Science, Springer, pp

    Anssi Yli-Jyrä, Masood Feyzbakhsh Rankooh & Tomi Janhunen (2023):Pruning Redundancy in Answer Set Optimization Applied to Preventive Maintenance Scheduling. In:PADL 2023, Lecture Notes in Computer Science, Springer, pp. 279–294, doi:10.1007/978-3-031-24841-2_18. 16Streamliners for Answer Set Programming A Encodings This section illustrates the encodings u...

    work page doi:10.1007/978-3-031-24841-2_18 2023