REVIEW 2 major objections 5 minor 48 references
LeanCSP: A Framework for Certifying Constraint Reformulation and Solving in Lean
T0 review · 2 major / 5 minor · reviewed 2026-07-31 · grok-4.5
Pith's one-line read A Lean framework proves constraint reformulations once for whole problem families and checks external solvers so you get end-to-end (un)satisfiability theorems without trusting the solver.
desk verdict Shipped Lean pipeline that actually attaches certificates to high-level CSPs and parametric reformulations—not just low-level formulas—with honest TCB and real SBC payoff numbers. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
Parametric π-equivalence and domain/variable symmetry-breaking constraints in Lean, combined with a library of 40+ constraints whose CSP-to-PB encoding is proved sound, so a checked VeriPB certificate against the in-Lean PB model implies unsatisfiability of the original CSP.
What would settle it
Exhibit a library CSP whose Lean model is accepted and certified unsatisfiable (or satisfiable) but that disagrees with the standard mathematical statement of the same combinatorial problem, or a soundness gap in the CSP-to-PB theorems that lets a satisfiable CSP produce a checked UNSAT certificate.
Extended reading notes
Core claim
LeanCSP is presented as the first end-to-end formally verified pipeline for a general constraint language: a CSP written in Lean is taken as the problem specification, reformulation properties (equivalence, equisatisfiability, symmetry-breaking soundness) are proved parametrically for whole families, and instance certificates from external solvers are checked back against that CSP (via a sound CSP-to-pseudo-Boolean translation and PBLean for UNSAT), so neither the translation nor the solver need be trusted.
Load-bearing premise
The CSP written in Lean is treated as the true mathematical specification of the problem; if that model is wrong about the intended question, the certificates only prove facts about the model.
Editorial extensions
If this is right
- A single parametric symmetry or equivalence proof can be reused for every instance size, turning one-time proof effort into repeated solver speedups (measured up to ~2×10^7 in deterministic operations).
- Combinatorial bounds such as S(4)=44, R(3,3)≤6, and W(2,3)≤9 can be stated as ordinary Lean theorems whose proofs go through external search plus in-Lean certificate checking.
- Unsound reformulations that pass testing on small n (e.g. a reversal leader on ordinary Schur that only fails at n=13) are refused because the required family-level symmetry proof cannot be given.
- Full in-Lean certification of the largest reported instances stays within a few minutes, so the trusted check is practical relative to solving.
Reading between the lines
- The same pattern—parametric reformulation proofs plus reflected certificate checking—could be applied to other CP global constraints or to optimization (not only decision) if dominance proofs are wired in the same way.
- Because the Lean CSP is the specification, library growth and modeling discipline become the main remaining sources of error; community libraries of verified encodings would amplify the approach.
- Choosing pseudo-Boolean certificates over SAT was driven by compact cutting-planes proofs on counting problems; families outside that regime may still prefer a SAT/LRAT path once fully wired.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper introduces LeanCSP, a Lean 4 framework that addresses two verification needs in constraint programming: (i) parametric proofs of reformulation properties—equivalence via π-equivalence, equisatisfiability, and soundness of variable/domain symmetry-breaking constraints—for entire CSP families built from a library of 40+ constraints; and (ii) instance-level certification of external solver answers by translating Lean CSPs to MiniZinc, SMT-LIB, or OPB, then checking SAT witnesses directly and UNSAT VeriPB proofs via PBLean against a soundness-proved CSP-to-PB encoding. Combining both layers yields end-to-end Lean theorems (e.g., schur_4_exact establishing S(4)=44) without trusting solvers or unverified translations. Experiments compare PB vs SAT proof scaling, measure verified SBC speedups up to ~2×10^7 in deterministic operations, and show in-Lean certification costs of at most a few minutes on the largest instances.
Significance. If the development is as claimed, this is a substantial systems contribution to trustworthy combinatorial solving: a general, library-uniform end-to-end pipeline inside a modern ITP, with parametric family-level reformulation proofs rather than per-instance certificates alone. Strengths that should count explicitly include the shipped Lean development and public code, machine-checked CSP-to-PB soundness for the library, the cautionary Schur reversal example that the framework refuses rather than silently accepts, concrete Lean theorems for classical bounds (S(4), R(3,3), W(2,3)), honest experimental reporting (including cases where SBCs do not help), and clear disclosure that the Lean CSP is taken as specification and that PBLean reflection adds Lean.ofReduceBool/Lean.trustCompiler to the TCB. The work cleanly separates formulation-level guarantees from solver certificates and is a natural foundation for further certified CP/SMT work in Lean.
major comments (2)
- [Section 3] Section 3 (TCB and specification paragraph): The central pipeline claim is sound as stated, but the manuscript should state more prominently—ideally in the abstract or introduction, not only mid-Section 3—that end-to-end theorems are theorems about the Lean CSP model taken as specification, and that the trusted base includes the Lean kernel plus the two reflection/compiler axioms used by PBLean. This is standard for the genre and already disclosed, but it is load-bearing for how readers will cite the S(4)=44-style results; a short explicit TCB box or paragraph would prevent over-reading the guarantees as theorems about an independent mathematical ground truth.
- [Section 3] Section 3 and library scope: UNSAT lifting rests on per-constraint CSP-to-PB soundness for the fixed library. The paper should state more sharply what happens when a user adds a new DynamicConstraint outside that library (no automatic UNSAT path until a new soundness lemma is proved), and whether the MiniZinc/SMT SAT path still applies unchanged. This does not break the main claim for library CSPs, but it bounds the “uniformly for every CSP built from our library” slogan and should be explicit so the contribution is not overstated.
minor comments (5)
- [Table 1, Table 2] Table 1 and Table 2: Consider adding a one-line note that Equiv./SB line counts are human proof effort (not generated), and that geo. mean speedups exclude Schur’s unsolved-without-SBC largest point as already footnoted—make the exclusion rule visible in the table caption itself.
- [Figure 2, Section 4] Figure 2: The odd-cycle panel supports the “gap is resolution, not encoding” reading; a sentence in §4 stating the CNF encoding used for the SAT baseline (and whether cardinality was encoded naively) would make the comparison fully reproducible from the text alone.
- [Section 1] Related work: A slightly sharper contrast with Dubois (2020) on verified non-binary-to-binary transformations and with VeriPB dominance/symmetry logging (Bogaerts et al., 2023) would help readers place the parametric Lean proofs versus instance-level proof logging; the distinction is present but could be one tighter paragraph.
- Presentation: The submitted text has many run-together words (likely PDF extraction artifacts, e.g., “formulation-levelproperties”, “2×10 7”). Ensure the camera-ready source has clean spacing and that code listings match the repository identifiers exactly.
- [Example 4] Example 4 / schur_4_exact: Mention certificate sizes and regeneration path is good; adding the Lean theorem statement’s dependency on the equisatisfiability lemma name in the text would help readers navigate the development.
Circularity Check
No significant circularity: end-to-end Lean certificates and parametric reformulation proofs do not reduce to fitted inputs or load-bearing self-citation chains.
full rationale
LeanCSP’s central claims are systems and formalization claims: parametric equivalence/equisatisfiability/SBC soundness in Lean, a library-wide CSP-to-PB soundness theorem, and instance-level certificate checking that yields Lean (un)satisfiability theorems about the stated CSP. Combinatorial demos (e.g. schur_4_exact, R(3,3) and W(2,3) bounds) re-establish classical external facts via that pipeline; they are not quantities defined from the method’s own outputs. Experimental speedups compare the same solver with/without verified SBCs on deterministic operation counts and wall time—empirical baselines, not fitted-then-predicted parameters. Citations to PBLean and related Lean tooling are infrastructure dependencies with independent checker semantics (reflection of a once-proved checker); they do not smuggle an ansatz or uniqueness theorem that forces the paper’s results by construction. The disclosed trust boundary (Lean CSP as specification; Lean.ofReduceBool/Lean.trustCompiler) is standard TCB disclosure, not circular derivation. No self-definitional loop, fitted-input-as-prediction, or renaming of a known empirical law appears in the derivation chain.
Assumptions & free parameters
assumptions (5)
- domain assumption Lean kernel correctness plus reflection axioms Lean.ofReduceBool and Lean.trustCompiler (Lean compiler in TCB) when PBLean/bv_decide-style checkers succeed
- domain assumption The Lean CSP model is identified with the mathematical problem specification (not merely an encoding of a separate ground truth)
- domain assumption For every library constraint, satisfiability of the CSP constraint implies satisfiability of its intermediate pseudo-Boolean representation (CSP→PB soundness)
- standard math Standard CSP solution/assignment semantics and classical definitions of equisatisfiability and solution-set bijection equivalence
- domain assumption Finite integer domains and constraints drawn from the predefined library for the translation/certification path
invented entities (2)
-
LeanCSP DynamicConstraint / dependent-type CSP encoding and π-equivalence layer
independent evidence
-
Generic value-precedence symmetry-breaking constraint proved once for interchangeable values
independent evidence
Cite this review
Pith. "Pith review of LeanCSP: A Framework for Certifying Constraint Reformulation and Solving in Lean." pith.science (2026). https://pith.science/paper/M344AYBG
@misc{pith2026260728459,
author = {Pith},
title = {Pith review of: LeanCSP: A Framework for Certifying Constraint Reformulation and Solving in Lean},
year = {2026},
howpublished = {\url{https://pith.science/paper/M344AYBG}},
note = {Machine review of arXiv:2607.28459}
}
read the original abstract
Constraint programming is a core technology for solving complex combinatorial problems in scheduling, planning, configuration, and verification. Trusting its results therefore demands guarantees at two levels: that reformulations applied beforehand are semantics-preserving, and that solvers produce correct answers. In this work, we introduce a framework that addresses both verification levels in the Lean theorem prover: it can be used to prove formulation-level properties, such as equivalence, equisatisfiability, and the correctness of symmetry-breaking constraints, parametrically for entire problem families; and to check solver-produced certificates for individual instances via translation backends to external formats such as MiniZinc, SMT-LIB, and OPB. Combining both levels yields an end-to-end workflow that establishes the satisfiability or unsatisfiability of a constraint problem without trusting the external solver. Experimental results show that our framework's verified symmetry breaking also pays off in practice: a single parametric proof per problem family, reused across all instance sizes, reduces solver search effort by a factor of up to 2x10^7, while the entire in-Lean certification stays affordable, taking at most a few minutes for our largest instances.
Figures
Reference graph
Works this paper leans on
-
[1]
M. Abdulaziz and F. Kurz. Formally Verified SAT-Based AI Planning . Proceedings of the AAAI Conference on Artificial Intelligence, 37 0 (12): 0 14665--14673, June 2023. ISSN 2374-3468. doi:10.1609/aaai.v37i12.26714
-
[2]
M. Alekhnovich. Mutilated chessboard problem is exponentially hard for resolution. Theoretical Computer Science, 310 0 (1-3): 0 513--525, Jan. 2004. ISSN 03043975. doi:10.1016/S0304-3975(03)00395-5
-
[3]
Barrett, A
C. Barrett, A. Stump, and C. Tinelli. The SMT-LIB Standard : Version 2.0. Technical report, Department of Computer Science, The University of Iowa, 2010
2010
-
[4]
Barrett, L
C. Barrett, L. de Moura, and P. Fontaine. Proofs in Satisfiability Modulo Theories . In All about Proofs , Proofs for All , volume 55 of Mathematical Logic and Foundations , pages 23--44. College Publications, London, UK, Jan. 2015. ISBN 978-1-84890-166-7
2015
-
[5]
A. Biere, T. Faller, K. Fazekas, M. Fleury, N. Froleyks, and F. Pollitt. CaDiCaL 2.0. In A. Gurfinkel and V. Ganesh, editors, Computer Aided Verification , volume 14681, pages 133--152. Springer Nature Switzerland, Cham, 2024. ISBN 978-3-031-65626-2 978-3-031-65627-9. doi:10.1007/978-3-031-65627-9_7
-
[6]
B. Bogaerts, S. Gocht, C. McCreesh, and J. Nordstr \"o m. Certified Dominance and Symmetry Breaking for Combinatorial Optimisation . Journal of Artificial Intelligence Research, 77: 0 1539--1589, Aug. 2023. ISSN 1076-9757. doi:10.1613/jair.1.14296
-
[7]
H. B \"o ving, S. Bhat, L. Cicolini, A. Keizer, L. Frenot, A. Mohamed, L. Stefanesco, H. Khan, J. Clune, C. Barrett, and T. Grosser. Interactive Bitvector Reasoning using Verified Bit-Blasting . Proceedings of the ACM on Programming Languages, 9 0 (OOPSLA2): 0 3259--3285, Oct. 2025. ISSN 2475-1421. doi:10.1145/3763167
doi:10.1145/3763167 2025
-
[8]
R. E. Bryant, A. Biere, and M. J. H. Heule. Clausal Proofs for Pseudo-Boolean Reasoning . In D. Fisman and G. Rosu, editors, Tools and Algorithms for the Construction and Analysis of Systems , volume 13243, pages 443--461. Springer International Publishing, Cham, 2022. ISBN 978-3-030-99523-2 978-3-030-99524-9. doi:10.1007/978-3-030-99524-9_25
Show all 48 references
-
[9]
Carlier, C
M. Carlier, C. Dubois, and A. Gotlieb. A Certified Constraint Solver over Finite Domains . In D. Giannakopoulou and D. M \'e ry, editors, FM 2012: Formal Methods , pages 116--131, Berlin, Heidelberg, 2012. Springer. ISBN 978-3-642-32759-9. doi:10.1007/978-3-642-32759-9_12
2012 doi
-
[10]
Carneiro
M. Carneiro. Lean4Lean : Verifying a Typechecker for Lean , in Lean . arXiv:2403.14064, Sept. 2025
2025 arXiv
-
[11]
G. Chu, P. J. Stuckey, A. Schutt, T. Ehlers, G. Gange, and K. Francis. Chuffed, a lazy clause generation solver. https://github.com/chuffed/chuffed, 2023
2023
-
[12]
C. R. Codel, J. Avigad, and M. J. H. Heule. Verified Encodings for SAT Solvers . In 2023 Formal Methods in Computer-Aided Design ( FMCAD ) , pages 141--151, Oct. 2023. doi:10.34727/2023/isbn.978-3-85448-060-0_22
2023 doi
-
[13]
Cohen, P
D. Cohen, P. Jeavons, C. Jefferson, K. E. Petrie, and B. M. Smith. Symmetry Definitions for Constraint Satisfaction Problems . In D. Hutchison, T. Kanade, J. Kittler, J. M. Kleinberg, F. Mattern, J. C. Mitchell, M. Naor, O. Nierstrasz, C. Pandu Rangan, B. Steffen, M. Sudan, D....
2005
-
[14]
W. Cook, C. Coullard, and Gy . Tur \'a n. On the complexity of cutting-plane proofs. Discrete Applied Mathematics, 18 0 (1): 0 25--38, Sept. 1987. ISSN 0166218X. doi:10.1016/0166-218X(87)90039-4
1987 doi
-
[15]
Cruz-Filipe , M
L. Cruz-Filipe , M. J. H. Heule, W. A. Hunt, M. Kaufmann, and P. Schneider-Kamp . Efficient Certified RAT Verification . In L. de Moura , editor, Automated Deduction -- CADE 26 , pages 220--236, Cham, 2017. Springer International Publishing. ISBN 978-3-319-63046-5. doi:10.1007...
2017 doi
-
[16]
Cruz-Filipe , J
L. Cruz-Filipe , J. Marques-Silva , and P. Schneider-Kamp . Formally Verifying the Solution to the Boolean Pythagorean Triples Problem . Journal of Automated Reasoning, 63 0 (3): 0 695--722, Oct. 2019. ISSN 1573-0670. doi:10.1007/s10817-018-9490-4
2019 doi
-
[17]
de Moura and S
L. de Moura and S. Ullrich. The Lean 4 Theorem Prover and Programming Language . In A. Platzer and G. Sutcliffe, editors, Automated Deduction -- CADE 28 , pages 625--635, Cham, 2021. Springer International Publishing. ISBN 978-3-030-79876-5. doi:10.1007/978-3-030-79876-5_37
2021 doi
-
[18]
R. Dechter. Constraint Processing . Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, Apr. 2003. ISBN 978-1-55860-890-0
2003
-
[19]
C. Dubois. Formally Verified Transformation of Non-binary Constraints into Binary Constraints . arXiv:2009.00583, Sept. 2020
2009 arXiv
-
[20]
Elffers and J
J. Elffers and J. Nordstr \"o m. Divide and conquer: Towards faster Pseudo-Boolean solving. In J. Lang, editor, Proceedings of the 27th International Joint Conference on Artificial Intelligence, IJCAI 2018, pages 1291--1299. ijcai.org, 2018. doi:10.24963/ijcai.2018/180
2018 doi
-
[21]
Flippo, K
M. Flippo, K. Sidorov, I. Marijnissen, J. Smits, and E. Demirovi \'c . A Multi-Stage Proof Logging Framework to Certify the Correctness of CP Solvers . In P. Shaw, editor, 30th International Conference on Principles and Practice of Constraint Programming ( CP 2024) , volume 30...
2024 doi
-
[22]
I. P. Gent, K. E. Petrie, and J.-F. Puget. Symmetry in Constraint Programming . In F. Rossi, P. van Beek, and T. Walsh, editors, Handbook of Constraint Programming , volume 2 of Foundations of Artificial Intelligence , pages 329--376. Elsevier, 2006. doi:10.1016/S1574-6526(06)80014-3
2006 doi
-
[23]
Gocht and J
S. Gocht and J. Nordstr \"o m. Certifying Parity Reasoning Efficiently Using Pseudo-Boolean Proofs . Proceedings of the AAAI Conference on Artificial Intelligence, 35 0 (5): 0 3768--3777, May 2021. ISSN 2374-3468. doi:10.1609/aaai.v35i5.16494
2021 doi
-
[24]
Gocht, C
S. Gocht, C. McCreesh, and J. Nordstr \"o m. An Auditable Constraint Programming Solver . In C. Solnon, editor, 28th International Conference on Principles and Practice of Constraint Programming ( CP 2022) , volume 235 of Leibniz International Proceedings in Informatics ( LIPI...
2022 doi
-
[25]
Gocht, C
S. Gocht, C. McCreesh, M. O. Myreen, J. Nordstr \"o m, A. Oertel, and Y. K. Tan. End-to- End Verification for Subgraph Solving . Proceedings of the AAAI Conference on Artificial Intelligence, 38 0 (8): 0 8038--8047, Mar. 2024. ISSN 2374-3468, 2159-5399. doi:10.1609/aaai.v38i8.28642
2024 doi
-
[26]
W. T. Gowers, B. Green, F. Manners, and T. Tao. On a conjecture of Marton . Annals of Mathematics, 201 0 (2): 0 515--549, Mar. 2025. ISSN 0003-486X, 1939-8980. doi:10.4007/annals.2025.201.2.5
2025 doi
-
[27]
A. Haken. The intractability of resolution. Theoretical Computer Science, 39: 0 297--308, 1985. ISSN 03043975. doi:10.1016/0304-3975(85)90144-6
1985 doi
-
[28]
M. J. Heule, W. A. Hunt, and N. Wetzler. Trimming while checking clausal proofs. In 2013 Formal Methods in Computer-Aided Design , pages 181--188, Oct. 2013. doi:10.1109/FMCAD.2013.6679408
2013
-
[29]
M. J. H. Heule. Schur number five. In 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 , AAAI '18...
2018 doi
-
[30]
M. N. Mansur, M. Christakis, V. W \"u stholz, and F. Zhang. Detecting critical bugs in SMT solvers using blackbox mutational fuzzing. In Proceedings of the 28th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineeri...
2020
-
[31]
Mohamed, T
A. Mohamed, T. Mascarenhas, H. Khan, H. Barbosa, A. Reynolds, Y. Qian, C. Tinelli, and C. W. Barrett. Lean- SMT : An SMT tactic for discharging proof goals in Lean . In Computer Aided Verification -- 37th International Conference, CAV 2025, Proceedings, Part III, volume 15933 ...
2025 doi
-
[32]
Nethercote, P
N. Nethercote, P. J. Stuckey, R. Becket, S. Brand, G. J. Duck, and G. Tack. MiniZinc : Towards a Standard CP Modelling Language . In C. Bessi \`e re, editor, Principles and Practice of Constraint Programming -- CP 2007 , pages 529--543, Berlin, Heidelberg, 2007. Springer. ISBN...
2007 doi
-
[33]
Nipkow, L
T. Nipkow, L. C. Paulson, and M. Wenzel. Isabelle/ HOL : A Proof Assistant for Higher-Order Logic . Number 2283 in Lecture Notes in Computer Science. Springer, Berlin New York, 2002. ISBN 978-3-540-45949-1
2002
-
[34]
Rossi, P
F. Rossi, P. van Beek , and T. Walsh, editors. Handbook of Constraint Programming. Foundations of Artificial Intelligence. Elsevier, 2006. ISBN 978-0-444-52726-4
2006
-
[35]
Roussel and V
O. Roussel and V. Manquinho. Input/ Output Format and Solver Requirements for the Competitions of Pseudo-Boolean Solvers . https://www.cril.univ-artois.fr/PB16/format.pdf, 2016
2016
-
[36]
Schulte, G
C. Schulte, G. Tack, and M. Z. Lagerkvist. Modeling and Programming with Gecode . https://github.com/Gecode/gecode, 2019. Corresponds to Gecode 6.2.0
2019
-
[37]
A. Stump. Verified Functional Programming in Agda . Number 9 in ACM Books. Association for Computing Machinery, New York, NY, 2016. ISBN 978-1-970001-27-3 978-1-970001-26-6 978-1-970001-25-9. doi:10.1145/2841316
2016 doi
-
[38]
Subercaseaux, W
B. Subercaseaux, W. Nawrocki, J. Gallicchio, C. Codel, M. Carneiro, and M. J. H. Heule. Formal Verification of the Empty Hexagon Number . LIPIcs, Volume 309, ITP 2024, 309: 0 35:1--35:19, 2024. ISSN 1868-8969. doi:10.4230/LIPICS.ITP.2024.35
2024 doi
-
[39]
S. Szeider. Leanback: MCP server for Lean 4 theorem proving --- check, prove, goals, eval, search. https://pypi.org/project/leanback/, 2026 a
2026
-
[40]
S. Szeider. PBLean : Pseudo-Boolean Proof Certificates for Lean 4. In 17th International Workshop on Pragmatics of SAT ( PoS 2026), a workshop of SAT 2026 and FLoC 2026 , Lisbon, Portugal, July 2026 b . To appear. Preprint available at https://arxiv.org/abs/2602.08692
2026
-
[41]
Colibrics: A Formally Verified Constraint Programming Engine
The Colibri Team . Colibrics: A Formally Verified Constraint Programming Engine . https://colibri.frama-c.com/, 2019
2019
-
[42]
The Lean Language Reference
The Lean Developers . The Lean Language Reference . https://lean-lang.org/doc/reference/latest/, 2026
2026
-
[43]
The lean mathematical library
The mathlib Community . The lean mathematical library. In Proceedings of the 9th ACM SIGPLAN International Conference on Certified Programs and Proofs , CPP 2020, pages 367--381, New York, NY, USA, Jan. 2020. Association for Computing Machinery. ISBN 978-1-4503-7097-4. doi:10....
2020
-
[44]
The Rocq Prover
The Rocq Development Team . The Rocq Prover . Zenodo, Apr. 2025
2025
-
[45]
E. Tsang. Foundations of Constraint Satisfaction. Academic Press, London San Diego, 1993. ISBN 978-0-12-701610-8
1993
-
[46]
van Doorn , P
F. van Doorn , P. Massot, and O. Nash. Formalising the h- Principle and Sphere Eversion . In Proceedings of the 12th ACM SIGPLAN International Conference on Certified Programs and Proofs , CPP 2023, pages 121--134, New York, NY, USA, Jan. 2023. Association for Computing Machin...
2023
-
[47]
Vanroose, I
W. Vanroose, I. Bleukx, J. Devriendt, D. Tsouros, H. Verhaeghe, and T. Guns. Mutational Fuzz Testing for Constraint Modeling Systems . In P. Shaw, editor, 30th International Conference on Principles and Practice of Constraint Programming ( CP 2024) , volume 307 of Leibniz Inte...
2024 doi
-
[48]
Wetzler, M
N. Wetzler, M. J. H. Heule, and W. A. Hunt. DRAT-trim : Efficient Checking and Trimming Using Expressive Clausal Proofs . In C. Sinz and U. Egly, editors, Theory and Applications of Satisfiability Testing -- SAT 2014 , pages 422--429, Cham, 2014. Springer International Publish...
2014 doi
Reviewed July 31, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.