Bundled Karp Reductions

Updated 24 January 2026

Bundled Karp reductions are a formalized approach that encapsulates the reduction function, explicit polynomial time bound, and correctness proof into one verifiable object.
They enable modular composition and mechanized verification in proof assistants, improving transparency in resource bounds and correctness analyses.
This method underpins automated hardness proofs in complexity theory and cryptography, ensuring precise transfer of algorithmic properties between problems.

A bundled Karp reduction is a formalized, data-structured approach to Karp (polynomial-time many-one) reductions in computational complexity. Unlike classical definitions which treat reductions as mere functions, bundled Karp reductions explicitly package the reduction function, its polynomial time complexity certificate, and correctness proof into a first-class mathematical object. This encapsulation is amenable to formal verification in proof assistants, compositional reasoning, and template-based reduction construction for both NP- and coNP-completeness arguments, as exemplified in recent formalization frameworks.

1. Formal Definition and Data Structure

A bundled Karp reduction between decision problems $(X, P_X)$ and $(Y, P_Y)$ consists of a computable mapping $f: X \to Y$ , a polynomial time witness $w: X \to \mathbb N$ , a proof of correctness

$\forall x \in X,\quad x \in P_X \;\Longleftrightarrow\; f(x) \in P_Y,$

and a formally-checkable polynomial bound

$\exists\,p\in\mathbb N[n],\ \forall x\in X:\ \text{runtime}(f,x)\leq p(|x|),$

where $|x|$ is the encoding length of $x$ and the bound is witnessed by $w(x)$ . This "bundling" enables not only explicit tracking of resource bounds but also mechanization of reductions and their composition in proof assistants such as Lean 4 (Simas, 22 Jan 2026).

In Lean, the structure is encapsulated as:

structure PolyReduction (X Y : Type _) [Encodable X] [Encodable Y] where
  reduce    : X → Y
  correct   : ∀ x, (x ∈ X.language) ↔ (reduce x ∈ Y.language)
  polyBound : ∃ (p : ℕ → ℕ), ∀ x, (time (reduce) x) ≤ p (sizeOf (encode x))

This approach enforces explicit proofs of membership preservation and polynomial-time computability, overcoming potential ambiguities of informally described reductions.

2. Motivations and Methodological Innovations

Bundled Karp reductions arise within efforts to formalize computational complexity and completeness theory, particularly as a response to the need for effective, scalable, and checkable construction of reductions within mechanized frameworks (Simas, 22 Jan 2026). The key methodological advances include:

Explicit polynomial runtime witnesses: Recording not just the existence but an explicit bounding function, supporting verified resource analyses.
First-class proofs and compositionality: Each reduction carries both a correctness equivalence and a time bound, which are preserved under generic composition operations.
Template-driven reduction frameworks: By uniformly packaging the data, automated tactics can build complex reductions (e.g., for $\Sigma_2^P$ - or coNP-completeness) from simpler components.

This formalization addresses the gap between textbook proofs (often omitting time analysis or assuming correctness) and the needs of complexity theory in proof verification systems.

3. Composition and Transitivity of Bundled Reductions

A fundamental property of bundled Karp reductions is compositionality: if $f$ is a bundled Karp reduction from $A$ to $B$ and $g$ from $B$ to $C$ , then their composition $h(x) = g(f(x))$ yields a bundled Karp reduction from $A$ to $C$ . The new polynomial time bound is $p_h(n) = p_g(p_f(n))$ , explicitly constructed from the bounds associated to $f$ and $g$ . This is formalized in Lean as:

1
2
3

theorem PolyReduction.comp_exists {A B C}
  (f : PolyReduction A B) (g : PolyReduction B C) :
  ∃ h : PolyReduction A C, ∀ a, h.reduce a = g.reduce (f.reduce a) := ...

Such composition reliably chains reductions, ensuring both correctness and explicit computational efficiency are maintained.

4. Applications in Formal Complexity Theory

Bundled Karp reductions have been used in recent formalizations to establish hardness of diverse problems, including the coNP-completeness of SUFFICIENCY-CHECK and $\Sigma_2^P$ -completeness for variants such as ANCHOR-SUFFICIENCY (Simas, 22 Jan 2026).

The methodology enables:

Uniform reduction construction: Each concrete reduction (such as from TAUTOLOGY to SUFFICIENCY-CHECK) is defined once as a PolyReduction object with explicit runtime analysis, then automatically participates in correctness and hardness proofs via compositional lemmas.
Separation of concerns: Boolean-circuit encodings, combinatorial gadget construction, and resource bounds are modularized.
Verified hardness proofs: By leveraging reusable templates, the proof of completeness is checked not only for correctness but for resource bounds as well.

The approach scales to chains involving code equivalence problems, as in bundled many-one reductions among PCE, LCE, SPCE, and LIP (Cheraghchi et al., 11 Feb 2025).

5. Significance for Complexity and Cryptography

By formalizing the data associated with a reduction, bundled Karp reductions clarify how computational hardness and algorithmic upper bounds transfer between problems. Their importance is marked in:

Practical hardness transfer: In code-based cryptography, polynomial-time chains of bundled reductions ensure that improvements or lower bounds for lattice isomorphism transfer to permutation or signed code equivalence problems (Cheraghchi et al., 11 Feb 2025).
Minimizing parameter blow-up: As with linear kernels for Karp's 21 NP-complete problems (Filar et al., 2019), explicit tracking of instance size plays a critical role in preserving the practical impact of reductions for optimization, cryptanalysis, and lower-bound theory.
Mechanized complexity proof verification: Bundled reductions are central to recent advances in formalized complexity theory, enabling robust, end-to-end verification of completeness and hardness results.

While classical Karp reductions suffice for most pen-and-paper proofs of NP-completeness, their formal verification requires additional bookkeeping for resource bounds and correctness. Bundled Karp reductions systematize this, supporting both manual and automated theorem proving.

A plausible implication is that widespread adoption of bundled reductions will standardize formal frameworks in complexity theory and enhance their applicability to cryptographic reductions, kernelization, and parameterized complexity. However, the creation and verification of explicit runtime bounds may impose additional proof burden compared to informal reductions, potentially requiring further automation in proof assistants.

7. Outlook and Research Directions

Continued development of bundled reduction frameworks, especially with libraries of reusable reduction templates and compositional tactics, is expected to underpin both new complexity-theoretic formalizations and cross-domain applications such as cryptographic security proofs and algorithm design. The ongoing formalization of complexity classes, hardness hypotheses, and isomorphism problems using bundled reductions is an active line of research (Simas, 22 Jan 2026), with the potential to further algorithmic meta-theorems and certified lower bounds.