An And-Sum Circuit with Signed Edges That Is More Succinct than SDD
Ryoma Onaka, Kengo Nakamura, Masaaki Nishino, Norihito Yasuda
Abstract
Knowledge compilation is a method of transforming knowledge into a compressed and tractable form for permitting more efficient operations. For Boolean functions, numerous representations have been proposed that enhance succinctness and tractability. In this paper, we introduce a new representation named structured Decomposable And-Sum Circuit (st-DASC), which employs AND and SUM nodes with signed edges, in place of the standard AND and OR nodes with unsigned edges. Notably, incorporating negative signs permits polytime logical negation. By following a knowledge compilation map, we show that st-DASCs are more succinct than Sentential Decision Diagrams (SDDs) while maintaining support for every operation on the knowledge compilation map that SDD supports. Furthermore, st-DASCs are even more succinct than structured d-DNNFs (st-d-DNNFs), which are more succinct than SDDs although they support fewer operations than SDDs. Accordingly, st-DASCs break the traditional trade-off between succinctness and tractability over SDDs and st-d-DNNFs.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext b5e1db26-5071-436b-935c-dced9ae4a8e6Cited by top-tier papers1
Ask how each one uses itBuilds on2
- Semantic Probabilistic Layers for Neuro-Symbolic LearningKareem Ahmed, Stefano Teso, Kai-Wei Chang, Guy Van den Broeck et al.NeurIPS 2022 · 133 citations
- A Compositional Atlas of Tractable Circuit Operations for Probabilistic InferenceAntonio Vergari, YooJung Choi, Anji Liu, Stefano Teso et al.NeurIPS 2021 · 112 citations
Related papers
- A Compiler for Weak Decomposable Negation Normal FormPetr Illner, Petr KuceraAAAI 2024
- Tensor Decomposition Meets Knowledge Compilation: A Study Comparing Tensor Trains with OBDDsRyoma Onaka, Kengo Nakamura, Masaaki Nishino, Norihito YasudaAAAI 2025 · 2 citations
- Lower Bounds on Intermediate Results in Bottom-Up Knowledge CompilationAlexis de Colnet, Stefan MengelAAAI 2022 · 2 citations
- Tractable Explanations for d-DNNF ClassifiersXuanxiang Huang, Yacine Izza, Alexey Ignatiev, Martin C. Cooper et al.AAAI 2022 · 43 citations
- Backdoor Decomposable Monotone Circuits and Propagation Complete EncodingsPetr Kucera, Petr SavickýAAAI 2021 · 1 citation
