---
title: Induced Interpolant Existence Problem
url: https://www.emergentmind.com/topics/induced-interpolant-existence-problem
type: topic
---

# Induced Interpolant Existence Problem

The induced interpolant existence problem is the instance-wise decision problem of whether, for a given entailment or contradiction, there exists an interpolant restricted to the shared vocabulary. In its standard form, given formulas \(P\) and \(Q\) in a logic \(L\), the problem asks whether there exists \(I \in L\) such that \(\Sigma(I)\subseteq \Sigma(P)\cap\Sigma(Q)\), \(P \models_L I\), and \(I \models_L Q\). In nonlinear real arithmetic, the same idea appears as the search for an interpolant \(I(x)\) over shared variables alone, typically of the form \(h(x)>0\), separating contradictory formulas \(\phi(x,y)\) and \(\psi(x,z)\) [2404.02683] [1903.01297]. When a logic has the Craig Interpolation Property (CIP), the problem collapses to validity; when CIP fails, interpolant existence becomes a separate semantic and algorithmic question, with decidability, complexity, and constructive behavior varying sharply across logical and algebraic settings [2404.02683] [2312.05929].

## 1. Formal formulation and principal variants

The core definition is uniform across much of the literature. For formulas \(A\) and \(B\) with shared signature \(S\), a Craig interpolant is a formula \(I\) such that \(A \Rightarrow I\), \(I \wedge B\) is unsatisfiable, and \(I\) mentions only symbols from \(S\). Equivalently, in entailment form, \(A \models I\) and \(I \models B\) with \(\Sigma(I)\subseteq \Sigma(A)\cap\Sigma(B)\) [1903.01297] [2404.02683].

A central distinction is between global interpolation and instance-wise existence. CIP asserts that every valid implication has such an interpolant. The interpolant existence problem asks only whether a specific implication does. This distinction is decisive in logics that do not enjoy CIP: in those settings, validity no longer guarantees interpolant existence, and the induced existence problem can be harder than entailment or even undecidable [2404.02683] [2312.05929] [2303.04598].

Several specialized variants recur. In nonlinear polynomial interpolation over the reals, one considers contradictory formulas
\[
\phi(x,y): f_1(x,y)\ge 0 \wedge \cdots \wedge f_m(x,y)\ge 0,\qquad
\psi(x,z): g_1(x,z)\ge 0 \wedge \cdots \wedge g_n(x,z)\ge 0,
\]
with shared variables \(x\), and seeks an induced interpolant \(I(x)\), usually \(h(x)>0\), such that \(\phi \Rightarrow h(x)>0\) and \(h(x)>0 \wedge \psi\) is unsatisfiable [1903.01297]. In proof-theoretic work, an “induced interpolant” can also mean an interpolant extracted directly from a proof object such as a resolution refutation, a clausal tableau, or an SMT proof tree [1705.05309] [2008.03489] [2508.11449].

Verification-oriented work generalizes the single-interpolant question to structured collections. Path interpolation, generalized simultaneous abstraction, state-transition interpolation, and tree interpolation ask whether one can obtain interpolants satisfying additional inductiveness or compositionality constraints across a sequence or tree of partitions [1212.4650].

## 2. Semantic characterizations of existence and nonexistence

A dominant pattern in modal, guarded, and description-logical settings is characterization by shared-signature indistinguishability. In \(\mathrm{FO}^2(E_1,E_2)\) and guarded fragments with two equivalence relations, the failure of interpolation is captured by a Robinson-style criterion: \(\phi\) and \(\psi\) have no interpolant iff there exist pointed models \(A,a\) and \(B,b\) such that \(A \models \phi(a)\), \(B \models \neg\psi(b)\), and \(A,a\) and \(B,b\) are bisimilar over the shared signature \(\rho\) [2404.02683]. Closely related criteria appear for modal logics above \(K4.3\), where no interpolant exists iff there are descriptive-frame models whose roots satisfy \(\varphi\) and \(\neg\psi\) while being \(\sigma\)-bisimilar, with \(\sigma = Var(\varphi)\cap Var(\psi)\) [2312.05929]. Description logics with nominals and role inclusions use the same pattern in the form of “joint consistency modulo \(L(\Sigma)\)-bisimulations,” and first-order modal fragments such as \(Q^1S5\) and \(S5_{\mathcal{ALC}^u}\) use \(\sigma\)-bisimulation consistency as the criterion for nonexistence [2007.02736] [2303.04598].

In nonlinear real arithmetic, the semantic picture is geometric rather than bisimulation-based. Under Archimedean assumptions on the quadratic modules generated by the constraints of \(\phi\) and \(\psi\), the \(x\)-projections
\[
P_x(\phi)=\{x\mid \exists y.\,\phi(x,y)\},\qquad P_x(\psi)=\{x\mid \exists z.\,\psi(x,z)\}
\]
are compact semialgebraic sets. If \(\phi \wedge \psi\) is unsatisfiable, then these projections are disjoint, and there exists a polynomial \(h(x)\) such that \(h>0\) on \(P_x(\phi)\) and \(h<0\) on \(P_x(\psi)\); consequently, \(h(x)>0\) is an interpolant [1903.01297]. For unbounded domains, homogenization and compactification replace the Archimedean boundedness hypothesis. If the homogenized closures \(\tilde{T}_1\) and \(\tilde{T}_2\) of the projected semialgebraic sets are disjoint, then there exists a semialgebraic interpolant of the form
\[
h(x)=h_1(x)+\sqrt{1+\|x\|^2}\,h_2(x),
\]
which strictly separates the two sets [2407.00625].

These characterizations divide the subject into two broad regimes. In relational logics, nonexistence is typically witnessed by shared-signature bisimilar or equivalent models. In semialgebraic settings, existence is typically witnessed by a separating function obtained from positivity certificates on projected feasible sets.

## 3. Decidability and complexity landscape

The complexity of induced interpolant existence is highly nonuniform.

| Setting | IEP status | Notes |
|---|---:|---|
| Finitely axiomatizable \(L \supseteq K4.3\) | coNP-complete | Same complexity as entailment [2312.05929] |
| Difference logic (DL) | coNP-complete | Polynomial-size bisimilar witnesses [2403.11255] |
| Weak \(K4\) | in coN3ExpTime; coNExpTime-hard | Validity is PSpace-complete [2403.11255] |
| \(\mathrm{FO}^2(E_1,E_2)\) | undecidable | Validity remains decidable [2404.02683] |
| \(GF^2\) with constants, \(E_1,E_2\) | undecidable | Also for corresponding DL counterparts [2404.02683] |
| \(Q^1S5\), \(S5_{\mathcal{ALC}^u}\) | in coN2ExpTime; 2ExpTime-hard | UI existence undecidable [2303.04598] |
| \(Q^1K\) | non-elementary decidable | UI existence undecidable [2303.04598] |
| DLs with nominals and/or role inclusions | 2ExpTime-complete with ontologies | Ontology-free/RI-only cases vary between coNExpTime-complete and 2ExpTime-complete [2007.02736] |
| LTL over finite words | decidable in 4ExpTime; PSPACE-hard | Via FO-separation of regular languages [2508.12805] |

Two global conclusions recur. First, interpolant existence need not track entailment complexity. Weak \(K4\) has PSpace-complete validity but an interpolant existence problem in coN3ExpTime with coNExpTime-hardness [2403.11255]. Description logics with nominals and role inclusions exhibit the same phenomenon: entailment is easier than existence, and the failure of CIP or projective Beth definability prevents direct reduction to validity [2007.02736]. Second, decidability of the base logic does not imply decidability of IEP. The first explicit counterexamples are \(\mathrm{FO}^2(E_1,E_2)\) and \(GF^2\) with constants and two equivalence relations, where satisfiability and entailment remain decidable but interpolant existence is undecidable [2404.02683].

There are also positive “same-complexity” zones. For finitely axiomatizable modal logics above \(K4.3\), IEP is coNP-complete and “not harder than entailment,” due to quasi-polysize bisimilar countermodels on descriptive frames [2312.05929]. The chapter-level treatment of interpolation and separation further shows that LTL IEP can be reduced to FO-separation for regular languages, yielding decidability in 4ExpTime over finite words and extending the semantic perspective from logic to language separation [2508.12805].

## 4. Constructive synthesis methods

In nonlinear polynomial interpolation over the reals, constructive existence is obtained from Putinar’s Positivstellensatz. Under Archimedean assumptions, one searches for \(h(x)\) and SOS multipliers \(u_i(x,y)\), \(v_j(x,z)\) satisfying
\[
h(x)-1 = u_{m+1}(x,y)+\sum_{i=1}^m u_i(x,y)f_i(x,y),
\]
\[
-h(x)-1 = v_{n+1}(x,z)+\sum_{j=1}^n v_j(x,z)g_j(x,z),
\]
with all multipliers SOS. These identities imply \(\phi \Rightarrow h\ge 1>0\) and \(\psi \Rightarrow h\le -1<0\), so \(h(x)>0\) is an interpolant. SOS constraints are encoded by Gram matrices and coefficient matching, yielding a feasibility SDP; under the Archimedean hypothesis, the method is complete for sufficiently large degree bounds. The same framework extends to disjunctions of conjunctions of polynomial inequalities [1903.01297].

The same paper addresses numerical soundness. Approximate SDP solutions are validated by PSD certification using floating-point Cholesky factorization with padding \(2\beta I\), together with coefficient error bounds \(\epsilon\). Over compact feasible regions, one derives certified inequalities
\[
\phi \Rightarrow h(x)\ge 1-\gamma_1\epsilon-\gamma_2\beta,\qquad
\psi \Rightarrow -h(x)\ge 1-\gamma_1\epsilon-\gamma_2\beta.
\]
If \(\gamma_1\epsilon < 1/2\) and \(\gamma_2\beta < 1/2\), then strict separation holds exactly [1903.01297].

For unbounded semialgebraic domains, homogenization moves the problem to a compact sphere:
\[
\tilde{f}(x_0,x)=x_0^{\deg(f)}f(x/x_0),\qquad x_0^2+\|x\|^2=1,\quad x_0\ge 0.
\]
Polynomial separation on the homogenized sets induces a semialgebraic interpolant
\[
h(x)=h_1(x)+\sqrt{1+\|x\|^2}\,h_2(x),
\]
and the paper provides complete SOS characterizations for both polynomial and semialgebraic interpolants, again reducible to SDP [2407.00625]. This strictly extends the polynomial case, because \(h_2 \equiv 0\) recovers polynomial interpolants.

Proof-based constructive methods cover SMT, arrays, propositional logic, and first-order tableaux. Proof Tree Preserving Interpolation handles mixed literals in SMT proofs for quantifier-free EUF plus linear arithmetic by introducing shared auxiliary variables, keeping the proof tree unchanged, and using specialized mixed-pivot rules for equalities and inequalities [1705.05309]. For the quantifier-free theory of arrays, Weak Equivalence on Arrays and PTPI yield sequence and tree interpolants from a single unsatisfiability proof, avoiding partition-dependent reruns of the solver; interpolants have worst-case exponential size for extensionality lemmas and worst-case quadratic size otherwise [1804.07173]. Clausal first-order tableaux support a two-stage construction: a bottom-up extraction of a ground interpolant \(\mathrm{ipol}(N_0)\), followed by interpolant lifting that replaces certain terms by quantified variables ordered by subterm dependency [2008.03489]. In classical propositional logic, quantifier elimination, DNF projection, resolution extraction, and tableau extraction all compute interpolants, with existence equivalent to unsatisfiability of \(A\wedge \neg B\) [2508.11449].

Hybrid modal logics add a further constructive result: for \(H\), \(H(@)\), \(H(@,U)\), and graded variants, hypermosaic elimination computes Craig interpolants, if they exist, in fourfold exponential time, with fourfold exponential output size [2602.15821].

## 5. Structured interpolants in verification

Verification workflows rarely need only a single interpolant. They require collections satisfying inductive or compositional constraints. A uniform framework organizes the principal collectives as path interpolation (PI), generalized simultaneous abstraction (GSA), state-transition interpolation (STI), and tree interpolation (TI) [1212.4650].

For an unsatisfiable sequence \(\phi_1,\dots,\phi_n\), PI requires interpolants \(I_0,\dots,I_{n-1}\) such that each prefix entails the corresponding \(I_k\), each \(I_k\) refutes the suffix, and variables are restricted to the prefix/suffix intersection. STI strengthens this to local Hoare-style steps
\[
(I_{\phi_1\cdots \phi_i,S_i}\wedge I_{\phi_{i+1},T_{i+1}})\models I_{\phi_1\cdots \phi_{i+1},S_{i+1}},
\]
while TI assigns interpolants to tree cuts [1212.4650].

In resolution-based Labeled Interpolation Systems, these collectives admit exact characterizations. For single systems, \(BGSA \Leftrightarrow GSA \Leftrightarrow TI \Leftrightarrow STI\), while PI and SA are weaker. For families of systems, TI is the most general collective, STI implies both PI and SA, and STI is equivalent to local BGSA constraints on every relevant triple. Necessary and sufficient labeling constraints are given by \(CC_{BGSA}\) and \(CC_{nGSA}\). All single LISs satisfy PI, systems at least as strong as Pudlák satisfy BGSA and therefore all stronger collectives, and McMillan’s dual \(Itp_{M'}\) separates PI from BGSA by satisfying the former but not the latter [1212.4650].

This framework directly informs model checking. Predicate abstraction and CEGAR require PI sequences; simultaneous abstraction and summary generation require SA or GSA; IC3/PDR-like inductive sequences correspond to STI; recursive or DAG-shaped decompositions require TI. The array-interpolation work supplies concrete sequence and tree interpolants for memory-safety proofs in C programs from one global proof, illustrating how theory-specific interpolation machinery realizes these collective requirements in practice [1212.4650] [1804.07173].

## 6. Boundaries, failures, and open directions

A fundamental limitation is that decidable quantified fragments may still fail to admit interpolants in-fragment. The array property fragment is a sharp example. The formulas
\[
A := \forall i.\ \mathrm{select}(a,i) < \mathrm{select}(b,k),\qquad
B := \forall j.\ \neg(\mathrm{select}(a,l) < \mathrm{select}(b,j))
\]
belong to the alternation-free APF and are jointly unsatisfiable, but no interpolant over the shared vocabulary exists within APF. The natural candidates
\[
I_1 := \exists j.\ \forall i.\ \mathrm{select}(a,i) < \mathrm{select}(b,j),\qquad
I_2 := \forall i.\ \exists j.\ \mathrm{select}(a,i) < \mathrm{select}(b,j)
\]
require quantifier alternation and lie outside the fragment [1904.11381]. This shows that decidability and closure under negation do not imply closure under interpolation.

Uniform interpolation is even more fragile. In \(Q^1S5\) and \(S5_{\mathcal{ALC}^u}\), uniform interpolant existence is undecidable [2303.04598]. In hybrid modal logics, the situation is broader: for any logic containing \(H\) and contained in FO, it is undecidable whether a uniform \(L(\sigma)\)-interpolant exists for a given formula [2602.15821]. These results sharply separate ordinary instance-wise interpolation from forgetting-style constructions.

Nonlinear algebraic methods have their own boundaries. The polynomial SOS approach of nonlinear Craig interpolation relies on Archimedean quadratic modules; if compactness fails, there may be no polynomial strictly separating the projected feasible sets, and the relevant SOS certificates may not exist [1903.01297]. The unbounded-domain extension mitigates this by homogenization, but at the cost of semialgebraic rather than purely polynomial interpolants and potentially severe degree growth and SDP conditioning issues [2407.00625].

Several boundary questions remain open. For \(\mathrm{FO}^2\) with a single equivalence relation, the status of IEP is open; likewise for dropping equality from \(\mathrm{FO}^2(E_1,E_2)\), for \(GF^2\) with counting, and for \(ALCQIO\) [2404.02683]. Above \(K4.3\), finite axiomatizability is crucial for the coNP procedure based on canonical formulas, leaving non-finitely axiomatizable extensions unresolved [2312.05929]. For weak \(K4\), the gap between the coN3ExpTime upper bound and coNExpTime-hardness is open, and constructive interpolation procedures remain a separate challenge [2403.11255]. In hybrid modal logics, the existence of fourfold-exponential algorithms leaves a nontrivial gap to triply-exponential lower bounds on separator size in the presence of \(U\) [2602.15821].

Taken together, these results place the induced interpolant existence problem at the intersection of model theory, real algebraic geometry, proof complexity, and formal verification. Its central question is simple—whether a shared-language separator exists for a given entailment—but its answer depends delicately on the ambient logic, the available proof objects, and the geometry or bisimulation theory underlying the formulas in question.

Source: https://www.emergentmind.com/topics/induced-interpolant-existence-problem