• Aucun résultat trouvé

Some progress on finding short proofs quickly

N/A
N/A
Protected

Academic year: 2022

Partager "Some progress on finding short proofs quickly"

Copied!
45
0
0

Texte intégral

(1)

Some progress on finding short proofs quickly

Anupam Das

University of Bath

Bath, April 21st 2011

This talk is available athttp://people.bath.ac.uk/ad402/FindingShortProofs.pdf

(2)

Outline

I Propositional proof systems

I Proof Size vs. Proof Search

I Gentzen vs. Deep inference systems

I A result for deep inference systems

(3)

Propositional proof systems

I A (sound and complete) proof system is a polynomial-time function whose range is the set of propositional tautologies.

I An object in the domain of such a function is called a “proof”.

I Examples:

(4)

Propositional proof systems

I A (sound and complete) proof system is a polynomial-time function whose range is the set of propositional tautologies.

I An object in the domain of such a function is called a “proof”.

I Examples:

(5)

Propositional proof systems

I A (sound and complete) proof system is a polynomial-time function whose range is the set of propositional tautologies.

I An object in the domain of such a function is called a “proof”.

I Examples:

(6)

Proof size

QUESTION: What is the relationship between the size of a tautology and the size of its proof?

I How small can we make proofs?

I Can we find a proof system where all proofs have size polynomial in the size of the tautologies they prove? Such a proof system is called super.

Theorem (Cook, Reckhow)

A super proof system exists iff N P =co-N P.

So the existence of a super proof system seems unlikely.

(7)

Proof size

QUESTION: What is the relationship between the size of a tautology and the size of its proof?

I How small can we make proofs?

I Can we find a proof system where all proofs have size polynomial in the size of the tautologies they prove? Such a proof system is called super.

Theorem (Cook, Reckhow)

A super proof system exists iff N P =co-N P.

So the existence of a super proof system seems unlikely.

(8)

Proof size

QUESTION: What is the relationship between the size of a tautology and the size of its proof?

I How small can we make proofs?

I Can we find a proof system where all proofs have size polynomial in the size of the tautologies they prove? Such a proof system is called super.

Theorem (Cook, Reckhow)

A super proof system exists iff N P =co-N P.

So the existence of a super proof system seems unlikely.

(9)

Proof size

QUESTION: What is the relationship between the size of a tautology and the size of its proof?

I How small can we make proofs?

I Can we find a proof system where all proofs have size polynomial in the size of the tautologies they prove? Such a proof system is called super.

Theorem (Cook, Reckhow)

A super proof system exists iff N P =co-N P.

(10)

Proof search

I PROBLEM: Given a tautology, how do we find a proof of it?

I Proof systems tend to be “infinitary”, and there is often no terminating algorithm to find a proof of a tautology.

I Example. Frege: A A→B

−−−−−−−−−−−

B . Acould be any formula, from which there are infinitely many to choose.

I Rules like modus ponens are instances of cut, and pose a big problem for proof search.

(11)

Proof search

I PROBLEM: Given a tautology, how do we find a proof of it?

I Proof systems tend to be “infinitary”, and there is often no terminating algorithm to find a proof of a tautology.

I Example. Frege: A A→B

−−−−−−−−−−−

B . Acould be any formula, from which there are infinitely many to choose.

I Rules like modus ponens are instances of cut, and pose a big problem for proof search.

(12)

Proof search

I PROBLEM: Given a tautology, how do we find a proof of it?

I Proof systems tend to be “infinitary”, and there is often no terminating algorithm to find a proof of a tautology.

I Example. Frege: A A→B

−−−−−−−−−−−

B . Acould be any formula, from which there are infinitely many to choose.

I Rules like modus ponens are instances of cut, and pose a big problem for proof search.

(13)

Proof search

SOLUTION! Gentzen’s Hauptsatz shows that we can eliminate cuts.

Results:

I Finitary proof search.

I Consistency of proof system.

I Subformula property.

However cut-free Gentzen systems can only at best exponentiallysimulate Gentzen systems with cut.

(14)

Proof search

SOLUTION! Gentzen’s Hauptsatz shows that we can eliminate cuts.

Results:

I Finitary proof search.

I Consistency of proof system.

I Subformula property.

However cut-free Gentzen systems can only at best exponentiallysimulate Gentzen systems with cut.

(15)

Proof search

SOLUTION! Gentzen’s Hauptsatz shows that we can eliminate cuts.

Results:

I Finitary proof search.

I Consistency of proof system.

I Subformula property.

However cut-free Gentzen systems can only at best exponentiallysimulate Gentzen systems with cut.

(16)

Statman tautologies

S1 ≡(c1d1)[¯c11] ,

S2 ≡(c2d2)[(([¯c22]c1)([¯c22]d1))[¯c11]] , S3 ≡(c3d3)(([¯c33]c2)([¯c33]d2))

((([¯c33][¯c22])c1)(([¯c33][¯c22])d1)) [¯c11] .

. . .

Cut-free sequent calculus proofs are exponentialin the size of the tautology.

(17)

Statman tautologies

S1 ≡(c1d1)[¯c11] ,

S2 ≡(c2d2)[(([¯c22]c1)([¯c22]d1))[¯c11]] , S3 ≡(c3d3)(([¯c33]c2)([¯c33]d2))

((([¯c33][¯c22])c1)(([¯c33][¯c22])d1)) [¯c11] .

. . .

Cut-free sequent calculus proofs are exponentialin the size of the

(18)

Question

How well can we optimise a system for both proof search and proof size?

(19)

Question

How well can we optimise a system for both proof search and proof size?

(20)

Deep inference

1 Inference rules operatearbitrarily deep within formulae. ξ (A

−−

B )

2 The distinction between object and meta level is abolished. Proofs are top-down symmetric.

Properties:

I Systems with cut are as efficient as their “shallow” counterparts.

I Exponential speedup for cut-free systems.

I There is a quasipolynomial cut-elimination procedure.

I Only has a restricted version of the subformula property.

Conjecture

All cuts can be eliminated in polynomial-time.

(21)

Deep inference

1 Inference rules operatearbitrarily deep within formulae. ξ (A

−−

B )

2 The distinction between object and meta level is abolished. Proofs are top-down symmetric.

Properties:

I Systems with cut are as efficient as their “shallow” counterparts.

I Exponential speedup for cut-free systems.

I There is a quasipolynomial cut-elimination procedure.

I Only has a restricted version of the subformula property.

Conjecture

All cuts can be eliminated in polynomial-time.

(22)
(23)
(24)
(25)
(26)
(27)
(28)

Deep inference

1 Inference rules operatearbitrarily deep within formulae. ξ (A

−−

B )

2 The distinction between object and meta level is abolished. Proofs are top-down symmetric.

Properties:

I Systems with cut are as efficient as their “shallow” counterparts.

I Exponential speedup for cut-free systems.

I There is a quasipolynomial cut-elimination procedure.

I Only has a restricted version of the subformula property.

Conjecture

All cuts can be eliminated in polynomial-time.

(29)

Deep inference

1 Inference rules operatearbitrarily deep within formulae. ξ (A

−−

B )

2 The distinction between object and meta level is abolished. Proofs are top-down symmetric.

Properties:

I Systems with cut are as efficient as their “shallow” counterparts.

I Exponential speedup for cut-free systems.

I There is a quasipolynomial cut-elimination procedure.

I Only has a restricted version of the subformula property.

(30)

Deep inference

I If conjecture is true then we have “finitary proof-search” with as short proofs as Frege/Gentzen with cut.

I While proof-search is now finitely branching, the “branching degree” is still unbounded, since we can now operate on any connective. Can we do better?

I QUESTION: How deep do we need to go?

(31)

Deep inference

I If conjecture is true then we have “finitary proof-search” with as short proofs as Frege/Gentzen with cut.

I While proof-search is now finitely branching, the “branching degree”

is still unbounded, since we can now operate on any connective. Can we do better?

I QUESTION: How deep do we need to go?

(32)

Deep inference

I If conjecture is true then we have “finitary proof-search” with as short proofs as Frege/Gentzen with cut.

I While proof-search is now finitely branching, the “branching degree”

is still unbounded, since we can now operate on any connective. Can we do better?

I QUESTION: How deep do we need to go?

(33)

Surprise!

Symmetry alone is enough! For propositional logic, the same speedup in proof size can be attained just from top-down symmetry in the system.

ξ{D}

ρ(2)−−−−−

ξ{D0} ρ0 :

ξ{D}

(A[B(CD)])

−−−−−−−−−−−−−−−−−−−−−−−−

(A[BC][BD])

ρ(1)−−−−−−−−−−−−−−−−−−−−−−−−−

(A[BC][BD0])

−−−−−−−−−−−−−−−−−−−−−−−−−

(A[B(CD0)])

ξ{D0}

(34)

Surprise!

Symmetry alone is enough! For propositional logic, the same speedup in proof size can be attained just from top-down symmetry in the system.

ξ{D}

ρ(2)−−−−−

ξ{D0} ρ0 :

ξ{D}

(A[B(CD)])

−−−−−−−−−−−−−−−−−−−−−−−−

(A[BC][BD])

ρ(1)−−−−−−−−−−−−−−−−−−−−−−−−−

(A[BC][BD0])

−−−−−−−−−−−−−−−−−−−−−−−−−

(A[B(CD0)])

(35)

Technicalities

I What about the equations?

Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(36)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(37)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws?

Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(38)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(39)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules?

Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(40)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(41)

Technicalities

I What about the equations? Turn them into inference rules and bound their depth too.

I Can we derive thedistributivity laws? Yes, in the presence of cocontraction.

I Can we still admit structural rules? Yes, and with no (significant) effect on proof complexity.

Theorem

Cut-free deep inference systems containing cocontraction can be restricted to allow just shallow inferences, with only polynomial increase in proof size.

(42)

So this means...

I Equates to augmenting a cut-free sequent system with the following two rules.

Γ, A∨B

−−−−−−−−

Γ, A, B

Γ, A∧B

−−−−−−−−−−−−

Γ, A Γ, B

I No longer a sequent calculus; it breaks the asymmetry induced by the tree-like structure of sequent calculus proofs.

I Less restricted version of subformula property than that for regular deep inference.

I Worse for proof search than Gentzen systems, but much better than regular deep inference while still giving access to short proofs.

(43)

So this means...

I Equates to augmenting a cut-free sequent system with the following two rules.

Γ, A∨B

−−−−−−−−

Γ, A, B

Γ, A∧B

−−−−−−−−−−−−

Γ, A Γ, B

I No longer a sequent calculus; it breaks the asymmetry induced by the tree-like structure of sequent calculus proofs.

I Less restricted version of subformula property than that for regular deep inference.

I Worse for proof search than Gentzen systems, but much better than regular deep inference while still giving access to short proofs.

(44)

Example

(45)

Thanks!

Références

Documents relatifs

tlaps is based on a hierarchical proof language that results in a clear proof structure and enables independent checking of the different proof obligations that arise in a proof..

Specific benefits include an XML proof format for TPTP proofs, links to provenance information maintained in the IW (e.g., information about ATP systems), structural search

The widespread increase in the level of diplomas, as the result of very different approaches to study on the part of students who sometimes continue their studies not out of

• Gain intuition towards finding lower bounds for normal proofs by studying the model theory of these arithmetics...

The transformation given induces a natural normalisation procedure to eliminate cocontractions in a proof, and we present some applications of this procedure where there is only

As introduc- tory material consult [BG09] and [Das12b] for known results on the complexity of deep inference, [Jeˇr09] for the relationship between deep inference and the mono-

3 we prove this to be false; we construct a polynomial transformation of cut-free deep inference proofs to ones whose inference rules are not only bounded in depth but “shallow”, in

However, the finer granularity of inference rules (one inference step in the sequent calculus corresponds to many inference steps in the calculus of structures) allows a finer