{Symmetry, Logic, Constraint Satisfaction Problem}
summary
Fixed-template finite-domain CSPs are hard exactly when they lack the right algebraic symmetries: existence of a nontrivial polymorphism clone (equivalently failure to admit a reduction from 3SAT via PP-interpretation) predicts tractability. Concrete tools: affine polymorphisms give small generating sets and Gaussian-elimination–style algorithms; Maltsev/majority polymorphisms yield polynomial local-generation and bounded width (singleton arc consistency suffices); absence of such identities yields NP-completeness. Extensions cover weighted polymorphisms for optimization, oligomorphic infinite templates, and open goals like an LP-unifying criterion and algebraic markers for undecidability.
AI-written summary · about
chapters (14)
- 0:00Intro
- 3:03CSP over fixed finite template
- 6:08Examples and a conjecture
- 8:07Selected results
- 11:17Example of simulation (gadget reduction, pp-definition)
- 13:001 reason for hardness
- 16:03Too popular viewpoint
- 17:45Alternative viewpoint
- 25:57Expressive power and polymorphisms
- 28:20Tractability conjecture again Tractability conjecture
- 29:39Tractability conjecture vs. reality
- 37:51Describing all solutions 2
- 43:59Beyond
- 48:57How far?
