---
title: Minimum Tangle Drop Overview
url: https://www.emergentmind.com/topics/minimum-tangle-drop
type: topic
---

# Minimum Tangle Drop Overview

Minimum tangle drop is not a single invariant across the literature. In current arXiv usage, the expression refers to several non-equivalent minimization or sensitivity quantities: a multipartite-entanglement monotone based on the least loss of one-to-group tangle under single-qubit tracing, a tanglegram crossing-number drop under deletion of one matching edge, and the minimum vertical extent of a wire tangle realizing a prescribed swap multiset. Closely related low-dimensional-topology usages quantify bridge-number loss under tangle product, minimum unknotting complexity under tangle replacement, and minimum untangling complexity for rational tangles [2606.21950], [1709.08119], [2312.16213], [1108.5640], [2410.13283], [1311.7171].

## 1. Scope of the term

The common pattern is an optimization over a local simplification or perturbation of a tangle-like object. What changes from field to field is the ambient category: quantum states, tanglegrams, layered adjacent-swap systems, or embedded tangles in \(3\)-manifolds. Only the multipartite-entanglement setting introduces a formal quantity explicitly built from a “minimum tangle drop”; elsewhere, the phrase is best understood as a convenient label for the minimum loss of some complexity measure [2606.21950].

| Domain | Quantity | Representative result |
|---|---|---|
| Multipartite entanglement | \(\mathcal D_{q_1}=\min_i \sqrt{\tau_{q_1|\mathrm{rest}}-\tau_{q_1|\mathrm{rest}\setminus q_i}}\) | LOCC monotone; for \(W_n\), \(\mathcal D_{\min}=2/n\) [2606.21950] |
| Tanglegrams | \(crt(T)-crt(T-e)\) | For size \(n\), the drop is at most \(n-3\), and this is sharp [1709.08119] |
| Wire tangles | Minimum number of layers realizing a swap list | NP-hard; exact exponential/FPT algorithms are known [2312.16213] |
| Tangle products | \(\beta(K_1)+\beta(K_2)-\beta(K_1 *_S K_2)\) | Under stated hypotheses, the drop is at most \(n(10n-6)\) [1108.5640] |

This divergence in usage matters methodologically. In the entanglement setting, the drop is a monotone derived from bipartite tangle. In combinatorics, it is an optimization target over layered realizations. In tanglegrams, it is a difference of crossing minima before and after edge deletion. In knot-theoretic settings, it is typically a lower-bounded simplification cost rather than a directly computable invariant.

## 2. Minimum tangle drop as a multipartite-entanglement monotone

The most explicit formal definition appears in multipartite entanglement. For an \(n\)-qubit state with focus qubit \(q_1\), the minimum tangle drop is defined by
\[
\mathcal{D}_{q_1}=\min_{i\in\{2,\dots,n\}}\sqrt{\tau_{q_1(q_2\dots q_n)}-\tau_{q_1(q_2\dots \hat q_i\dots q_n)}}.
\]
Here \(\tau_{q_1(q_2\dots q_n)}\) is the one-to-group tangle across the bipartition \(q_1\,|\,q_2\dots q_n\), and \(\tau_{q_1(q_2\dots \hat q_i\dots q_n)}\) is the corresponding tangle after tracing out \(q_i\). For pure \(2\times d\) states, the underlying tangle is
\[
\tau_{A(BC)}=2[1-\operatorname{Tr}(\rho_A^2)]=4\det(\rho_A),
\]
and for mixed states the construction is extended by convex roof [2606.21950].

This quantity measures the least degradation of the global entanglement of the focus qubit under loss of one other particle. It is therefore simultaneously a one-focus robustness diagnostic and a multipartite entanglement monotone. The cited work states that the convex-roof extension is LU-invariant, convex, and monotonic under LOCC, and that it vanishes on biseparable states. It also states that the quantity fails strict positivity on all genuinely multipartite entangled states, so it is not a faithful GME detector in full generality [2606.21950].

The same framework introduces a minimum negativity drop as a computationally efficient variant. That variant is motivated by the difficulty of mixed-state tangle evaluation after tracing out one qubit, and by the desire to retain non-vanishing values on \(W\)-class states. This suggests that “minimum drop” is best viewed as a family of loss-of-correlation probes, with the tangle-based version occupying the most direct conceptual position and the negativity-based version supplying computational tractability [2606.21950].

## 3. Tripartite reduction, exact families, and limitations

In the tripartite regime, the minimum tangle drop collapses to a more classical quantity. If one defines
\[
\mathcal D_{\min}=\min_{k\in\{1,2,3\}}\mathcal D_{q_k},
\]
then the cited work proves
\[
\mathcal D_{\min}
=
\sqrt{\tau_{ABC}+\min\{\tau_{AB},\tau_{AC},\tau_{BC}\}},
\]
so the symmetrized minimum tangle drop is physically equivalent to the minimum pairwise concurrence. This gives a particularly transparent three-qubit interpretation: the drop combines irreducible tripartite entanglement with the weakest residual two-qubit channel [2606.21950].

The framework yields exact formulas for \(W\)-type families. For the pure \(n\)-qubit \(W\) state,
\[
\mathcal D_{\min}(W_n)=\frac{2}{n},
\]
and for the noisy family
\[
\rho=(1-\gamma)|W_n\rangle\langle W_n|+\gamma|0\dots0\rangle\langle 0\dots0|,
\]
the corresponding value is
\[
\mathcal D_{\min}=\frac{2(1-\gamma)}{n}.
\]
These expressions exhibit both positivity on \(W\)-class states and \(1/n\)-type robustness scaling [2606.21950].

The same work uses drop profiles as structural probes of entanglement architecture. Nonzero drop under removal of \(x\) indicates that \(x\) belongs to a mutually inseparable subset attached to the focus qubit, and this supports cluster reconstruction heuristics and graph-state fingerprinting. At the same time, the paper gives explicit limitations. A chain-type graph state is presented as a genuinely multipartite entangled state with vanishing minimum entanglement drop, and the broader drop framework is said to lose diagnostic sensitivity on highly robust states such as the \(5\)-qubit error-correcting code. A plausible implication is that minimum tangle drop is strongest as a local vulnerability probe and weaker as a universal scalar classifier of multipartite entanglement [2606.21950].

## 4. Edge-deletion drop in tanglegrams

In tanglegram theory, the relevant notion is the decrease in minimum crossing number after deleting one matching edge. A tanglegram \(T=(L,R,M)\) consists of two rooted binary trees with the same number of leaves and a perfect matching \(M\) between their leaf sets. In a layout, the trees are drawn plane on opposite sides of a strip, their leaves lie on two parallel lines, and only matching edges may cross. The tangle crossing number \(crt(T)\) is the minimum number of crossing pairs of matching edges over all layouts [1709.08119].

For a matching edge \(e\), the drop is measured by
\[
crt(T)-crt(T-e),
\]
where \(T-e\) is obtained by deleting the endpoints of \(e\) and suppressing the resulting degree-\(2\) vertices. The central theorem states that if \(T\) has size \(n\ge 3\), then
\[
crt(T)-crt(T-e)\le n-3,
\]
and the bound is sharp. The proof starts from an optimal layout of \(T-e\), reinserts the deleted edge by local switch choices in the two trees, and shows that one can always restore \(e\) so that it crosses at most \(n-3\) other matching edges [1709.08119].

Sharpness is realized by an explicit family \(P_n\) of rooted-caterpillar tanglegrams. For each \(n\ge 4\), \(P_n\) is \(1\)-edge tangle planar, has
\[
crt(P_n)=n-3,
\]
and becomes planar after deleting a distinguished matching edge. Hence
\[
crt(P_n)-crt(P_n-e)=n-3.
\]
This shows that a single edge can account for all nonplanarity of a tanglegram while still forcing only a linear drop, even though the global extremal tangle crossing number is quadratic: if \(\gamma(n)\) denotes the maximum tangle crossing number over size-\(n\) tanglegrams, then
\[
\frac12\binom{n}{2}(1-o(1))\le \gamma(n)<\frac12\binom{n}{2}.
\]
The same paper also gives an \(O(n^4)\)-time algorithm for nontrivial lower bounds on \(crt(T)\), which is relevant when one wants to estimate deletion drops computationally [1709.08119].

A closely related 2025 result isolates the first nonplanar level. If a tanglegram has exactly one cross-responsible size-\(4\) subtanglegram, then its tangle crossing number is exactly \(1\). This does not define a “drop” quantity, but it gives a structural certificate for the minimum achievable nonzero crossing count [2505.00163].

## 5. Minimum vertical extent in wire tangles

In the combinatorics of adjacent swaps, the natural “drop” is the height of a tangle. A tangle consists of \(n\) \(y\)-monotone wires arranged on horizontal layers, with consecutive layers differing only by swaps of neighboring wires. A list \(L=(l_{ij})\) records how many times each pair of wires must swap. A tangle realizes \(L\) if the multiset union of swaps between consecutive layers equals \(L\); \(\textsc{List-Feasibility}\) asks whether such a tangle exists, and \(\textsc{Tangle-Height Minimization}\) asks for one with the minimum number of layers [2312.16213].

The paper proves that \(\textsc{List-Feasibility}\) is NP-complete even if every pair of wires swaps at most eight times. Since height minimization subsumes feasibility, the optimization problem is NP-hard as well. The positive side is an exact DP over sublists. If \(H(L')\) denotes the minimum height of a tangle realizing a sublist \(L'\), and \(\rho=id_nL'\) is the final permutation forced by parity, then
\[
H(L')=\min_{\pi\in S_\rho} H\bigl(L'-L(\langle \pi,\rho\rangle)\bigr)+1,
\]
where \(S_\rho\) is the set of permutations adjacent to \(\rho\). The resulting algorithm solves \(\textsc{Tangle-Height Minimization}\) in
\[
O\!\left(\left(2|L|/n^2+1\right)^{n^2/2}\varphi^n n\log|L|\right),
\]
with \(\varphi\approx 1.618\) arising from the Fibonacci count of adjacent predecessor permutations [2312.16213].

The same work derives a simpler feasibility-only DP with running time
\[
O\!\left(\left(2|L|/n^2+1\right)^{n^2/2}n^3\log|L|\right),
\]
shows that \(\textsc{List-Feasibility}\) is fixed-parameter tractable with respect to the number of wires \(n\), and gives an FPT algorithm running in
\[
O\big((n/2)^{n^2}\cdot n^3\log n + n^2\log |L| \big).
\]
For simple lists, where every swap occurs at most once, the minimum-height problem becomes cleaner: if \(L(\pi)\) denotes the inversion list of a permutation \(\pi\), then \(L(\pi)\) is the unique simple list with \(id_nL(\pi)=\pi\), and shortest-path search on the resulting permutation graph yields an
\[
O(n!\varphi^n)
\]
algorithm [2312.16213].

This line of work generalizes earlier height-optimality results. An earlier paper gave an exact algorithm with running time
\[
O\!\left(\left(2|L|/n^2+1\right)^{n^2/2}\cdot \varphi^n \cdot n\right)
\]
in the abstract formulation and treated simple lists in \(O(n!\varphi^n)\) time, but the later formulation also establishes NP-membership and fixed-parameter tractability with respect to the number of wires [1901.06548], [2312.16213].

## 6. Related minimum-complexity notions in knot and tangle theory

In low-dimensional topology, several adjacent notions measure how much complexity can be lost under a tangle operation. For \(n\)-strand tangle products \(K_1 *_S K_2\), one paper proves that if there exists a minimal bridge sphere for \(K_1 *_S K_2\) of distance at least three and the product sphere \(S\) is \(c\)-incompressible, then
\[
\beta(K_1 *_S K_2)\ge \beta(K_1)+\beta(K_2)-n(10n-6).
\]
Equivalently, if one defines the bridge-number drop by \(\beta(K_1)+\beta(K_2)-\beta(K_1 *_S K_2)\), then the drop is at most \(n(10n-6)\). The proof passes through a bound of \(10n-8\) on the number of saddles in the induced foliation of the \(2n\)-punctured product sphere [1108.5640].

A different Floer-theoretic notion measures the minimum local tangle complexity required to unknot a knot. If \(d_{ot}(K,U)\) denotes the minimum \(n\) such that an oriented \((n+1)\)-tangle replacement unknots \(K\), and \(d_t(K,U)\) the analogous unoriented quantity, then
\[
d_{ot}(K,U)\ge \operatorname{Ord}(K),\qquad d_t(K,U)\ge \operatorname{Ord}'(K),
\]
where \(\operatorname{Ord}(K)\) and \(\operatorname{Ord}'(K)\) are knot Floer torsion orders. More generally,
\[
d_{ot}(K,K')+\operatorname{Ord}(K')\ge \operatorname{Ord}(K).
\]
This makes torsion order a lower bound on minimum one-shot tangle-replacement simplification [2410.13283].

For rational tangles, the minimum untangling complexity is Euclidean-algorithmic. If
\[
x_{i-1}=x_ip_i+\epsilon_{i+1}x_{i+1}
\]
is a generalized Euclidean algorithm, then the corresponding untangling sequence has
\[
(m-1)+\sum_{i=1}^m p_i
\]
moves, where the moves are twists plus rotations. The paper proves that the method of least absolute remainders and the regular Euclidean algorithm have the same minimal number of such steps, and that this is the smallest possible. Thus the most efficient untangling of a rational tangle is determined by the minimal subtraction-style Euclidean step count [1311.7171].

Finally, the minimum crossing complexity of essential tangles is also known sharply in several regimes. For an essential \(n\)-string tangle, the minimum crossing number is at least \(2n+1\); if all strings are unknotted, the minimum rises to \(2n+2\); and for essential \(n\)-string \(k\)-loop tangles with \(k\ge 1\), the bound is \(2(n+k)-2\). These are minimum-complexity statements rather than “drop” statements in the strict sense, but they belong to the same family of extremal tangle simplification questions [1505.07268].

Across these literatures, minimum tangle drop is therefore best understood as a family resemblance rather than a single invariant. The shared theme is minimal loss or minimal simplification under a local operation, but the ambient measures—tangle, crossing number, height, bridge number, Floer torsion, or untangling steps—are domain-specific and mathematically distinct.

Source: https://www.emergentmind.com/topics/minimum-tangle-drop