Lemma 5.1 — in the gadget
ProvedBunkbedFalse.gadget_connConsider the gadget graph of Fig. 2: vertices and , where each spoke – is retained with probability and each path edge – is retained with probability . For and ,
The proof is a two-case recurrence on whether the spoke is open, giving with , solved by induction.
This is the first of three elementary computations that together determine how well the gadget simulates a single hyperedge, which is the quantitative heart of the refutation.
Formalization Note Vertices are Fin (n+1) with and , so is Fin.last n. The weight function assigns to any edge meeting and to the rest.
import Definitions.Def_BunkbedHypergraph open Bunkbed Bunkbed.Hyper SimpleGraph Finset
namespace BunkbedFalse
/-- **Lemma 5.1** (Gladkov–Pak–Zimin, PNAS 122 (2025) e2420725122, p. 8). For the gadget
graph `Gₙ` of Fig. 2, `ℙ_P(a ↔ vₙ) = (1 - P^{2n})/(1 + P)`. -/
theorem gadget_conn (n : ℕ) (P : ℚ) (hn : 3 ≤ n) (hP0 : 0 < P) (hP1 : P < 1) :
connProb (gadgetE n) (gadgetW n P) 0 (Fin.last n) = (1 - P ^ (2 * n)) / (1 + P) := by sorry
end BunkbedFalseRead-back
What the Lean code literally says, in plain math · claude-opus-5
Read-back — BunkbedFalse.gadget_conn
What the statement is about
The declaration fixes a natural number and a rational number , together with three hypotheses: , , and . It then asserts a closed-form value for a single Bernoulli-bond-percolation connection probability on one explicit finite graph.
The graph
The vertex set is ( vertices). The edge set consists of exactly two families:
- Spokes: for every — i.e. vertex is joined to every other vertex. ( edges.)
- Path edges: for every — a path running , not touching vertex . ( edges.)
So the graph is a fan/wheel-like graph on edges: a path on plus a hub adjacent to all path vertices. (Verified by evaluation: for the edge set is exactly .)
The weights
Each edge is retained independently, with retention probability
Note the spokes carry and the path edges carry , not the other way round. (Verified: at the three spokes of have weight and the two path edges have weight .)
The probability is defined in the fully elementary way: sum over all subsets of the edge set, weighting by , and keeping those for which the two distinguished vertices lie in the same component of the graph with vertex set and edge set .
The two vertices
The connection event is between:
- — the hub, the vertex incident to all spokes; and
- (written
Fin.last n) — the far endpoint of the path, which is also the unique path-endpoint of largest index. Note is itself adjacent to the hub via a spoke, so the two vertices are neighbours.
The claim
Since divides , the right-hand side is in fact the polynomial ; the identity is therefore an identity of polynomials in , not merely of numbers on .
Numerical verification
I built a scratch file evaluating both sides. The identity HOLDS at every tested point (exact rational equality):
| both sides | ||
|---|---|---|
| 3 | ||
| 3 | ||
| 3 | ||
| 3 | ||
| 3 | ||
| 4 | ||
| 4 | ||
| 4 | ||
| 4 | ||
| 5 | ||
| 5 | ||
| 5 | ||
| 5 |
(The value is a poor test point on its own, since there and the spoke/path weight distinction vanishes; the asymmetric values do distinguish them and all match.)
Hypotheses: used, needed, degenerate cases
- is stronger than needed, but is not vacuous. The identity also holds at ( at ) and ( at ). It fails at : there the graph has the single vertex , no edges, the left side is (a vertex is reachable from itself), and the right side is . So excludes the one failing , with excluded gratuitously.
- is not needed. At both sides are ; at both sides equal (for ) — the identity persists outside , where neither side is a probability.
- excludes exactly one genuinely bad point. At both sides are , so positivity is not needed there; but at the left side is (for ) while the right side is , because rational division by zero returns . So is doing work only at the single point .
- The claim is not trivial for any admissible or : under , both sides are strictly between and and depend on (at : , , for ).
Indexing remark
The pair of vertices is (hub, path-endpoint-). The closed form is specific to a path endpoint: at , the hub-to- probabilities are
and . So the formula matches the two path endpoints and (which are interchangeable by the path's reflection symmetry) and not the interior vertices. Also gives trivially.
Other literal features
- The declaration lives in namespace
BunkbedFalseand is stated withsorry(no proof). - It imports the hypergraph-percolation file, but every symbol it uses (, the weights, the connection probability) comes from the Bernoulli-percolation file; nothing from the hypergraph/Wierman–Ziff layer, and nothing about two-level "bunkbed" graphs, posts, or transversal sets, appears in the statement. The statement concerns a single-level graph only.
Confirmed by the mission captain (proposal self-audit).