Overlap-Determinability in Theory & Applications

• Overlap-Determinability is a domain-general concept that formalizes the computability, testability, and operational use of overlap between sets, states, or regions.
• It underpins frameworks such as CoDO in network analysis and quantifies overlap for effective causal inference and quantum superposition.
• The concept bridges disciplines—including theoretical computer science, quantum information, combinatorics, and experimental physics—ensuring precise criteria for overlap evaluation.

Overlap-Determinability is a domain-general concept capturing the conditions under which the presence, magnitude, or structure of overlap—between sets, states, or regions—can be reliably inferred, rigorously quantified, or operationally utilized in inference, measurement, or transformation. The notion arises across theoretical computer science, statistics (especially causal inference), network science, combinatorics, quantum information, and experimental physics, where the “determinability” of overlap entails either its explicit computability, testability, or the existence of protocols whose correctness hinges on well-specified overlap information.

1. Foundational Principles and Definitions

At its core, overlap-determinability formalizes whether and how overlap data can be unambiguously extracted, manipulated, or operationalized, with criteria specific to the problem domain:

• Statistical Overlap Determinability: The ability to test if observed overlap (e.g., between two subgraphs or subsets) is unexpectedly large under an appropriate null model, and to assign a statistically meaningful significance score (typically a $p$-value) to the event, as in the CoDO framework (Magner et al., 2016).
• Operational Overlap Determinability (Quantum Information Theory): The existence of a physical convention or side-information enabling the relative phase between quantum pure states—effectively making the (complex) overlap between their respective projectors a well-defined and operationally accessible quantity, thus enabling otherwise forbidden superposition procedures (Bang et al., 21 Jan 2026).
• Combinatorial and Algorithmic Determinability: The capacity to algorithmically construct, extend, or minimize overlap representations (e.g., in overlap graphs), or to decide properties about overlap-free/inclusive structures, with determinability quantified in both computational and structural senses (Rosgen et al., 2010, Schaeffer et al., 2022).
• Causal Inference and Experimental Design: The ascertainment of regions in covariate or policy space where overlap between groups (i.e., the positivity or “support” assumption) holds, thereby determining the interpretable or externally valid regions for effect estimation and transportability (Oberst et al., 2019, Huang, 2024).

2. Statistical and Algorithmic Quantification of Overlap

Statistical frameworks for overlap-determinability focus on assigning calibrated significance metrics to the observed overlap between sets, clusters, or subgraphs in data-rich domains:

• CoDO (Combining Density and Overlap): This framework computes the probability, under an Erdős–Rényi null model, that two random subgraphs of observed sizes would have (i) an overlap of size at least $z$ and (ii) an edge density within the overlap at least as large as observed, yielding the $p_{\rm CoDO}$ value. The calculation involves double hypergeometric summations, encoding both size and density information, and produces a significance measure monotonic in both overlap size and density (Magner et al., 2016).
• Theoretical Properties: CoDO provides large-deviation guarantees, smooth transition between the contributions of size and density, and is robust in distinguishing biologically or socially meaningful overlaps even when purely size- or density-based tests fail.
• Interpretability and Ranking: $p_{\rm CoDO}$ values are interpretable in the standard $p$-value sense, allowing ranking of overlaps, multiple testing correction, and threshold selection tailored for the application domain.

Empirically, CoDO distinguishes subtly overlapping but functionally relevant structures in protein–protein interaction networks and social circles better than previous approaches, with AUC and correlation metrics outperforming hypergeometric-size and pure-density tests (Magner et al., 2016).

3. Operational and Structural Overlap-Determinability in Quantum and Discrete Settings

Quantum Information: Phase Ambiguity and Superposition

The operational determinability of quantum state overlap is shown to be equivalent to the existence of a phase convention on the set of rays (projectors) comprising the domain of interest. The main theorem in (Bang et al., 21 Jan 2026) establishes that a probabilistic superposition device (CPTNI map) exists on a domain $\Omega$ if and only if $\Omega$ is overlap-determinable, i.e., a fixed phase convention can be established:

• Full versus Restricted Domains: For pairs of unknown states, universal overlap-determinability is provably impossible, ruling out universal superposition and related forbidden operations (cloning, deletion, masking, superluminal signaling).
• Structural Phase Liftings: When the domain is restricted (e.g., all states lying on a common submanifold or promised side information fixing Pancharatnam phases), a phase convention becomes operational, realizing overlap-determinability for that set (Bang et al., 21 Jan 2026).
• Computational Implications: The absence of overlap-determinability enforces the no-superposition and no-cloning theorems and secures the computational lower bound for Grover search. In contrast, universal access to overlap would collapse the quantum and communication complexity separation.

Combinatorics and Graph Representation

• Overlap Number in Graphs: For a graph $G$, the overlap number is the minimum cardinality of a universe needed such that sets assigned to each vertex overlap if and only if the vertices are adjacent, without one set containing another (Rosgen et al., 2010). The determinability question is whether, for a given $G$ and $k$, there exists an overlap representation of size $\leq k$.
• Algorithmic Complexity: Efficient (polynomial-time) determinability applies for classes such as cliques, paths, cycles, and caterpillars via closed formulas. However, the extension and containment-constrained decision problems are NP-complete, and the full overlap number problem is widely conjectured to be NP-hard.

Domain	Principle of Overlap-Determinability	Decidability/Complexity
Quantum superposition	Phase convention fixes overlap among rays	Exists if and only if said convention is supplied (Bang et al., 21 Jan 2026)
Causal inference	Regions identified by minimum-volume set and bounded propensity	Tractable via interpretable classifier; $O(1/\sqrt{n})$ estimation error (Oberst et al., 2019)
Network cluster overlap	Statistical significance via $p_{\rm CoDO}$ (size and density)	Efficient; polynomial in overlap size (typically $O(10^2)$) (Magner et al., 2016)
Graph overlap number	Minimal overlap representations subject to adjacency constraints	Efficient for special classes; NP-complete in general (Rosgen et al., 2010)

4. Overlap-Determinability in Causal Inference and External Validity

In observational and policy evaluation studies, overlap-determinability is instantiated as the capacity to localize regions in covariate ($X$) or policy space where all treatments (or actions) have bounded propensity ($\eta_t(x)\in[\epsilon,1-\epsilon]$). Failure of overlap (i.e., positivity violations) fundamentally limits the identifiability of causal effects for the corresponding regions:

• Characterization and Estimation: The intersection $O^{\alpha,\epsilon} = S^\alpha \cap B^\epsilon$ defines the region with sufficient support and bounded propensity. Learning interpretable overlap rules is reduced to a Neyman–Pearson-style volume-constrained classification (with DNF formulas).
• Off-Policy Generalization: For policy evaluation, overlap-determinability is further specialized to $B^\epsilon(\pi)$, guaranteeing sufficient representation for all actions under the target policy (Oberst et al., 2019).
• Quantitative Sensitivity: Sensitivity analysis using the decomposition $\text{Bias} = \delta\cdot\gamma$ partitions the bias from overlap violations into (i) the proportion of the population omitted ($\delta$) and (ii) the difference in average treatment effect for the omitted segment ($\gamma$) (Huang, 2024). Robustness and benchmarking tools support practical interpretability and empirical study design.

5. Determinability in Combinatorial Structures and Word Theory

The concept of overlap-determinability extends to discrete structures, such as sequences and their representations:

• Automata-Theoretic Decidability of Overlap-Free Properties: For infinite binary words, Schaeffer and Shallit (Schaeffer et al., 2022) show the first-order theory of overlap-free (or more generally, $\alpha$-power-free) words is decidable via finite automaton representations (DFAO/Büchi automata). Any first-order statement about such words, including existence of overlaps or power factors, reduces to automata emptiness, enabling algorithmic proofs for otherwise combinatorially complex properties.

6. Experimental Physics: Overlap-Determinability in Imaging

In particle imaging and velocimetry, overlap determinability is formalized through statistical models predicting the expected degree of overlap among particle images (PIs) in a field. The universal parameter is the non-dimensional seeding density $\mathcal{S}$, interpreted as the fraction of the image occupied by PIs:

• Overlap Laws: Key overlap statistics (mean overlaps per PI, fraction of overlap-free PIs, mean usable PI area) are closed-form functions of $\mathcal{S}$, validated experimentally across modalities and size distributions (Sax et al., 22 Jun 2025).
• Thresholds for Design: A critical threshold at $\mathcal{S}=0.25$ corresponds to each PI overlapping, on average, one other. Practical guidelines express the determinability of the overlap regime (acceptable, critical, or excessive) as a function of experimentally controllable parameters.

7. Practical and Theoretical Implications

Overlap-determinability serves as a unifying constraint or enabler across fields:

• In quantum information, it marks the boundary between physically permitted and forbidden transformations, preserving both foundational postulates and computational lower bounds (Bang et al., 21 Jan 2026).
• In network science and statistical inference, it operationalizes significance testing for overlap phenomena, facilitating hypothesis generation and cluster validation in high-dimensional data (Magner et al., 2016).
• In causal inference, it circumscribes the valid region for identification and generalization of treatment effects, demanding explicit reporting and sensitivity analysis regarding overlap violations (Oberst et al., 2019, Huang, 2024).
• In combinatorial optimization and discrete mathematics, it governs representability and decidability regimes for graph and word properties, interfacing automata theory with complexity theory (Rosgen et al., 2010, Schaeffer et al., 2022).

The field continues to expand with the development of frameworks and algorithms for reliably detecting, measuring, and leveraging overlap, and in clarifying the limitations imposed by the absence of overlap-determinability in both physical and abstract systems.