On Multiplicative Linear Logic, Modality and Quantum Circuits

A logical system derived from linear logic and called QMLL is introduced and shown able to capture all unitary quantum circuits. Conversely, any proof is shown to compute, through a concrete GoI interpretation, some quantum circuits. The system QMLL, which enjoys cut-elimination, is obtained by endowing multiplicative linear logic with a quantum modality.

💡 Research Summary

The paper introduces a novel logical system called QMLL (Quantum Multiplicative Linear Logic), which extends the traditional multiplicative fragment of linear logic (MLL) with a pair of quantum modalities. These modalities, denoted □ and ◇, are not merely modal operators in the usual sense; they are designed to encode unitary quantum operations directly within the logical syntax. The authors first formalize the syntax of QMLL, where formulas are built from atomic propositions (interpreted as qubits) using the tensor (⊗) and par (⅋) connectives of MLL together with the new modalities. □A is read as “prepare A as a quantum state,” while ◇A means “apply a quantum transformation to A.” This design ensures that each logical formula corresponds to a specific fragment of a quantum circuit.

The proof system of QMLL augments the standard MLL inference rules (axiom, cut, ⊗‑introduction, ⅋‑elimination) with three additional rules governing the quantum modalities: □‑introduction, ◇‑elimination, and a quantum cut rule. The quantum cut rule captures the composition of unitary operations when two modal formulas are connected, mirroring the sequential composition of gates in a circuit. Crucially, the authors prove a cut‑elimination theorem for QMLL, showing that any proof can be transformed into a cut‑free normal form while preserving provability. This property guarantees strong normalization and provides a logical counterpart to circuit simplification.

A central technical contribution is the adaptation of the Geometry of Interaction (GoI) semantics to the quantum setting. In classical GoI, a proof is interpreted as a token moving through a graph, with the token’s path encoding the dynamics of cut elimination. In QMLL, the token carries a complex‑valued state vector; when it traverses a □ or ◇ node, the corresponding unitary matrix is applied to the vector. The authors formalize this interaction as a linear operator on a Hilbert space and prove that the GoI interpretation of any QMLL proof yields exactly the unitary matrix described by the associated quantum circuit. Consequently, there is a one‑to‑one correspondence between proof structures and circuit diagrams.

The paper establishes two complementary completeness results. First, for any finite‑size unitary circuit C built from a universal gate set (e.g., Hadamard, CNOT, Phase), the authors present a constructive translation that builds a QMLL proof whose GoI semantics reproduces C. The translation proceeds gate‑by‑gate: each gate is encoded as a small proof fragment using the quantum modalities, and sequential composition is realized via the quantum cut rule. The size of the resulting proof grows linearly with the number of gates, demonstrating that QMLL is expressive enough to capture all unitary quantum computations. Second, they prove a soundness (or “negative”) property: any QMLL proof, when interpreted through the quantum GoI, necessarily yields a unitary operator. This follows from the fact that the modalities are defined only for unitary matrices and that the inference rules preserve unitarity. Hence, QMLL does not admit non‑unitary or measurement‑like operations, preserving the logical consistency of the system.

Beyond these core results, the authors discuss the meta‑theoretical implications of cut elimination for quantum circuit optimization. Removing cuts in a proof corresponds to eliminating redundant gate compositions, suggesting that proof‑theoretic techniques could be repurposed for circuit simplification and resource analysis. Moreover, because QMLL provides a formal logical language for describing quantum circuits, it opens the door to formal verification of quantum programs: one can reason about circuit equivalence, correctness, and resource usage within a well‑understood proof system.

In summary, the paper delivers a rigorous bridge between multiplicative linear logic and quantum circuit theory. By introducing quantum modalities and a quantum‑aware GoI semantics, it shows that QMLL both faithfully represents any unitary circuit (completeness) and only generates unitary transformations (soundness), while enjoying cut‑elimination and strong normalization. These results lay a solid foundation for future work on logic‑based quantum programming languages, formal verification tools for quantum algorithms, and deeper explorations of the structural parallels between proof theory and quantum computation.