1. What this unit proves 2
2. What it uses 11
3. Strategy 13
4. Steps 14
4.1 Algebra of the finite gas 14
4.2 Tree-graph bound 15
4.3 Sums over trees 15
4.4 Cluster sums and the logarithm 16
4.5 Deleted regions and forced polymers 16
4.6 Packets and the large-coupling numerics 17
4.7 The logarithm by deletion 19
4.8 The sector decomposition 21
4.9 Matrices 22
5. Where the difficulty lies 24
6. Notation 25
7. What remains unexplained 26
The full account of unit III.5: every proposition with its proof sketch, checked against the Lean in two rounds and edited. The condensed outline and the map of the Lean are on the unit page.
Unit page · All units · Overview
Part: III, Ultraviolet construction (renormalization group). Lean: work packages YM/Push/ABSOLUTE_SUMS/ (3,624 lines), YM/AnalyticNext/CONNECTED_ACTIVITY_BOUNDS/ (3,589), YM/Push/SECTOR_MEASURE/ (1,153), YM/SectorNext/VACUUM_NORMALIZATION/ (935), YM/Progress/STATIONARY_SCENES/ (445). Size: 9,746 lines in 1,097 declarations used by the main theorem.
Conventions.
\(\Lambda_n = (\mathbb{Z}/ n\mathbb{Z})^4\), with \(n \geqslant 1\), is the periodic lattice with the maximum distance \(d\). A support is a subset of \(\Lambda_n\).
\(\text{cov}_B (X)\) is the least number of \(d\)-balls of radius \(B\) whose union is connected and covers \(X\) (coverNumber). Two properties are used throughout (I.1):
\(|X| \leqslant \text{ball}_B \hspace{0.17em} \text{cov}_B (X)\), with \(\text{ball}_B = (2 B + 1)^4\);
\(\text{cov}_B\) is subadditive on unions of overlapping sets.
A family of supports is connected if it is nonempty and its overlap graph is connected.
For \(f \geqslant 0\) on an arbitrary index set, \(\text{SB} (f ; C)\) means: \(f\) is summable and \(\sum f \leqslant C\) (SumBudget).
For polymers \((X_i, w_i)_{i \in \iota}\) with \(w \geqslant 0\), the site-sum condition \(\text{Hit}_B (w ; a, \gamma)\) means \(\text{SB} \left( i \mapsto \boldsymbol{1}[x \in X_i] \hspace{0.17em} w_i \hspace{0.17em} e^{a \hspace{0.17em} \text{cov}_B (X_i)} ; \gamma \right)\) for every site \(x\) (Ac.Hit, II.2).
All weights are real. An upper bound on an infinite sum of nonnegative terms includes summability, as in \(\text{SB}\).
For a relation \(R\) that is not symmetric, the graph used below has an edge between distinct occurrences \(a, b\) exactly when both \(R (z_a, z_b)\) and \(R (z_b, z_a)\) hold. The coefficient of a one-vertex tuple is \(1\).
The unit consists of five packages.
(A) Mayer expansion.
The Mayer expansion of a hard-core polymer gas, with absolute bounds, and its re-exponentiation after forced polymers (III.5.1–III.5.6).
The verification of the smallness condition for gases of large-field seeds of a block-spin step (III.5.7–III.5.8).
(B) Vacuum normalization. A second, graph-free construction of the logarithm by deletion identities (III.5.9), and a trace-log formula (III.5.10).
(C) Sector decomposition. The decomposition of one block-spin integral into sectors labelled by packets of seeds, with a Gaussian gain per seed (III.5.11).
(D) Inverse decay. A Combes–Thomas bound (III.5.12).
(E) Gaussian bounds. Windowed inverses (III.5.13), and a uniform spectral gap and uniform decay for the precision matrix of the small-field Gaussian of a step (III.5.14).
III.5.1–III.5.7, III.5.9, III.5.10, III.5.12 and III.5.13 are general implications. III.5.8(c), III.5.11 and III.5.14 concern the block-spin step under the funding hypotheses of III.1; III.5.14 also fixes a patch satisfying its branch conditions.
Proposition III.5.1 (Mayer identity for a finite gas).
Data. Let \(\iota\) be finite, \(R\) a reflexive symmetric relation on \(\iota\) ("incompatible"), and \(w : \iota \to \mathbb{R}\). Let \(\Xi (w) = \sum_s \prod_{i \in s} w_i\), the sum over subsets \(s \subseteq \iota\) of pairwise compatible labels.
Ursell coefficient. For \(z = (z_0, \ldots, z_m) \in \iota^{m + 1}\), let \(G_R (z)\) be the graph on \(\{0, \ldots, m\}\) with an edge \(\{a, b\}\), \(a \neq b\), whenever \(R (z_a, z_b)\). Put \(\phi^T_R (z) = \sum_G (- 1)^{|G|}\), the sum over connected spanning subgraphs \(G \subseteq G_R (z)\).
Claim. Let \(\ell_m = \frac{1}{(m + 1) !} \sum_{z \in \iota^{m + 1}} \phi^T_R (z) \prod_a w_{z_a}\). If \(| \ell_m | \leqslant Cq^m\) for all \(m\), with \(C \geqslant 0\) and \(0 < q < 1\), then \(\exp \left( \sum_{m \geqslant 0} \ell_m \right) = \Xi (w)\).
Status: theorem. Lean: Gas.exact_polynomial; coefficient identities Coef.recurrence, Partition.partition.
Proposition III.5.2 (tree-graph bound). Let \(T\) be a graph on a finite vertex set with a distinguished vertex. Then \(| \sum_G (- 1)^{|G|} |\), summed over the connected spanning subgraphs \(G \subseteq T\), is at most the number of spanning trees of \(T\). In particular \(| \phi^T_R (z) |\) is at most the number of spanning trees of \(G_R (z)\). Status: theorem. Spanning trees are counted as acyclic parent maps towards the distinguished vertex. Lean: Graph.pay.
Proposition III.5.3 (pinned cluster bounds).
Data. \(\iota\) is any set; \(X : \iota \to\) supports of \(\Lambda_n\); \(B \in \mathbb{N}\); \(w : \iota \to [0, \infty)\); \(a \geqslant 0\); \(R\) is a relation on \(\iota\) with \(R (i, j) \Rightarrow X_i \cap X_j \neq \emptyset\). \(R\) need not be symmetric.
Smallness condition.
Notation. \(X (z) = \bigcup_a X_{z_a}\) for a tuple \(z\).
Then, for every site \(x\) and every integer \(m \geqslant 0\):
(a) \(\displaystyle \sum_{m' \geqslant m} \sum_{z \in \iota^{m' + 1} : x \in X (z)} \frac{| \phi^T_R (z) |}{(m' + 1) !} \prod_a w_{z_a} e^{a \hspace{0.17em} \text{cov}_B (X (z))} \leqslant \frac{\gamma \hspace{0.17em} q^m}{1 - q}\).
(b) The same bound holds for \(\sum_s \prod_{i \in s} \left( w_i e^{a \hspace{0.17em} \text{cov}_B (X_i)} \right)\). The sum runs over finite sets \(s \subseteq \iota\) with \(|s| \geqslant m + 1\) whose supports form a connected family containing \(x\).
(c) (Marked form.) Suppose \(Q \subseteq \iota\) and \(\text{SB} (i \mapsto \boldsymbol{1}[i \in Q] w_i e^{(a + \delta) \text{cov}_B (X_i)} ; b)\). Then the sum over \(z \in \iota^{m + 1}\) with some \(z_a \in Q\) of \(| \phi^T_R (z) | \prod_{j = 0}^m \left( w_{z_j} e^{a \hspace{0.17em} \text{cov}_B (X_{z_j})} \right) / (m + 1) !\) is at most \(bq^m\).
Each displayed bound is a summability statement of the form \(\text{SB}\). Status: theorem. Form (a) is used in III.2, III.10, III.11, III.12, III.13 and III.15. Lean: Ursell.log_hit_tail, Trees.touch_tail, Ursell.cluster_at, Ursell.cluster_tail; hypothesis Trees.Ready.
Proposition III.5.4 (convergence and logarithm of the gas). Let \(X, B\) be as in III.5.3, \(R\) reflexive and symmetric with \(R (i, j) \Rightarrow X_i \cap X_j \neq \emptyset\), \(w : \iota \to \mathbb{R}\) and \(a \geqslant 0\). Assume \(\text{Ready} (|w| ; a, \gamma, \delta, q)\). Then:
\(\Xi (w) = \sum_s \prod_{i \in s} w_i\), over finite pairwise compatible \(s \subseteq \iota\), converges absolutely, with \(\sum_s \prod_{i \in s} |w_i | \leqslant e^{n^4 \gamma}\).
The cluster series \(\sum_{m \geqslant 0} \sum_{z \in \iota^{m + 1}} \phi^T_R (z) \prod_a w_{z_a} / (m + 1) !\) converges absolutely. Its absolute sum is at most \(n^4 \gamma / (1 - q)\), and at most \(n^4 \gamma q^m\) over the clusters of \(m + 1\) labels.
\(\Xi (w) > 0\), and \(\log \Xi (w)\) equals the cluster series.
Status: theorem. Used in III.10. Lean: Gas.price_log, Gas.absolute, Gas.absolute_at.
Proposition III.5.5 (removing the polymers that meet a region). In III.5.4, take for \(R\) the overlap relation (\(R (i, j)\) iff \(i = j\) or \(X_i \cap X_j \neq \emptyset\)), with nonempty supports. For \(Y \subseteq \Lambda_n\) let \(w^Y_i = w_i \hspace{0.17em} \boldsymbol{1} [X_i \cap Y = \emptyset]\). Then
where \(p\) is the sum of the cluster terms with \(X (z) \cap Y \neq \emptyset\).
Hence \(e^{- C |Y|} \Xi (w) \leqslant \Xi (w^Y) \leqslant e^{C |Y|} \Xi (w)\) with \(C = \gamma / (1 - q)\), uniformly in \(n\).
The clusters of at least \(m + 1\) labels that meet \(Y\) contribute at most \(|Y| \gamma q^m / (1 - q)\) in absolute value.
Status: theorem. Used in III.10. Lean: Gas.restores_log, Gas.missing_tail.
Proposition III.5.6 (re-exponentiation with forced polymers). Fix \(B \in \mathbb{N}\) and a common lattice \(\Lambda_n\).
Data. "Unforced" polymers \((X_i, u_i)_{i \in \iota}\) and "forced" polymers \((Y_k, d_k)_{k \in \kappa}\), all with nonempty supports, \(u, d \geqslant 0\); \(R\) a relation on \(\iota\) contained in the overlap relation.
Atoms. The atoms are of two kinds:
the forced labels \(k\), with support \(Y_k\) and weight \(d_k\);
the tuples \(z \in \iota^{m + 1}\), with support \(X (z)\) and weight \(| \phi^T_R (z) | \prod_a u_{z_a} / (m + 1) !\).
Constants. \(A \geqslant 0\), \(\delta' > 0\), \(t \geqslant 1\), \(0 \leqslant \sigma < 1\), \(\mu \geqslant 0\), \(\eta > 0\), \(q' < 1\).
Hypotheses. \(\text{Ready} (u ; A + \delta', \gamma, \delta, q)\), \(\text{Hit}_B (d ; A + \delta', c)\), and
Fix a site \(x\) and \(m \geqslant 0\). Consider finite sets \(s\) of atoms with multiplicities \(n_{\alpha} \geqslant 1\) such that:
\(s\) contains at least one forced label;
\(\sum_{\alpha} n_{\alpha} \geqslant m\);
the supports of the atoms in \(s\) form a connected family containing \(x\).
Summed over all such pairs \((s, (n_{\alpha})_{\alpha \in s})\), the quantity
is at most \(t^{- m} e^{- \mu} \hspace{0.17em} \eta / (1 - q')\).
Status: theorem. The general form, with an arbitrary selection predicate, is Multiple.estimates. Used in III.2, III.10, III.11 and III.13. Lean: OmitClusters.restore.
Proposition III.5.7 (gases of seed packets).
Setting. Let \(S\) be a block-spin step (III.1) from \(\Lambda_f\) to \(\Lambda_c\) with block side \(L\), and let \(l, B \in \mathbb{N}\).
A seed is a fine cell, a coarse bond or a coarse site. Its location is the projection of the cell’s vertices, the bond’s two endpoints, or the site itself. Each location lies within distance \(1\) of its anchor.
A packet \(A\) (Stores) is a finite set of seeds; \(|A|\) is its number of seeds.
Let \(D_0 (A)\) be the union of its seed locations and \(D_r (A)\) its radius-\(r\) neighbourhood. The skins are the radius-\(lB\) neighbourhoods of the individual seed locations; their union is the dome \(D (A) = D_{lB} (A)\). The packet is connected if its skins form a connected family. The sector and matrix estimates below use \(D_r (A)\) directly, rather than neighbourhoods of \(D (A)\).
Put
(a) Packet sums. Assume \(a, C \geqslant 0\), \(0 \leqslant p_A \leqslant C \hspace{0.17em} e^{(s - g) |A|}\) for every connected \(A\), and \(\rho := K_{\text{step}}^2 \hspace{0.17em} e^{\hspace{0.17em} s - g + a (2 l + 3)^4} \leqslant q < 1\). Then for every coarse site \(x\) and every \(m \geqslant 0\),
The same bound holds for an arbitrary family of variants with nonnegative weights, each nonzero variant assigned to a connected packet and supported in its dome, provided the sum of the variant weights of each packet is at most \(Ce^{(s - g) |A|}\). For variants, the sum is pinned by the actual support and uses its covering number.
(b) Mixed words.
Alphabet. Add optional polymers \((X_i, w_i)_{i \in \iota}\) on \(\Lambda_c\) with \(w \geqslant 0\) and lengths \(\lambda_i \geqslant 0\). A packet has footprint \(D (A)\) and length \(e_s |A|\).
Hypotheses (with \(\varepsilon, \delta, \gamma_{\text{root}} > 0\) and \(a, C, e_s, \mu_s \geqslant 0\)):
\(0 \leqslant p_A \leqslant Ce^{(s - g) |A|}\) on connected packets;
(a) holds with \(a + 2 \varepsilon + \delta\) in place of \(a\), \(s + 2 \varepsilon e_s + \mu_s\) in place of \(s\), and some \(q_p < 1\);
\(\text{Hit}_B (w \hspace{0.17em} e^{2 \varepsilon \lambda} ; \hspace{0.17em} a + 2 \varepsilon + \delta, \hspace{0.17em} \gamma_{\text{opt}})\);
\(K_{\text{hit}} Cq_p / (1 - q_p) + \gamma_{\text{opt}} \leqslant \gamma_{\text{root}}\);
\(4 \hspace{0.17em} \text{ball}_B \hspace{0.17em} \gamma_{\text{root}} / \delta \leqslant q < 1\).
Notation. For a finite set \(s\) of packets and optional labels, \(D (s)\) is the union of the footprints, and \(\text{ch} (s) = \sum_{\alpha \in s} (\text{cov}_B (\text{footprint of } \alpha) + \text{length of } \alpha)\).
Claim. For all \(m, k, r \in \mathbb{N}\), \(T \in \mathbb{R}\) and every site \(x\),
the sum over connected \(s\) with \(x \in D (s)\), \(|s| \geqslant m + 1\), at least \(r\) seeds in total, and \(\text{ch} (s) \geqslant T\). Packets that are not connected have weight \(0\).
Status: theorem. The hypotheses of (b) are the records Mixed.Costs.Buy and Mixed.Accounts. (b) is used in III.2 and III.10–III.13. Lean: Packets.packet_income, Packets.packet_variants_tail, Mixed.mixed_hit, Mixed.ready_vertices.
Proposition III.5.8 (the numerical conditions hold at large coupling).
Notation. Fix the gauge group. In (a) and (b), also fix the block side \(L\) and \(l \in \mathbb{N}\). For the coarse coupling \(x\) put:
\(\beta_F = x + b \log L\), with \(b\) a constant of the group;
\(B = \lceil (\log x)^2 \rceil\);
\(P (s) = e^{(\log \log s)^2}\);
\(t = \beta_F^{- 1 / 2} P (\beta_F)\), the small-field threshold;
\(\text{bill} (x) = 1 + B + (\log \beta_F)^2\);
for \(\mu, u > 0\), \(M = \mu u \hspace{0.17em} P (\beta_F)^2\).
(a) Let \(a, D \geqslant 0\), \(d \in \mathbb{N}\) and \(z < 2\). For all sufficiently large \(x\) the following holds. Whenever \(s \leqslant D \hspace{0.17em} \text{bill}^d P (\beta_F)^z\), \(g \geqslant M\) and \(0 \leqslant C \leqslant \exp (D \hspace{0.17em} \text{bill}^d P (\beta_F)^z)\):
the numerical condition \(\rho \leqslant q < 1\) of III.5.7(a) holds with \(q = e^{- M / 2}\);
\(K_{\text{hit}} Cq^{\max (1, m)} / (1 - q) \leqslant e^{- M \max (1, m) / 4}\) for every \(m\).
If \(\beta \geqslant u \beta_F\), then \(\beta \mu t^2 \geqslant M\).
(b) Let in addition \(\varepsilon, \delta > 0\). For all large \(x\), all \(\beta \geqslant u \beta_F\), and all \(C, e_s \geqslant 0\), \(s \in \mathbb{R}\), and \(\gamma_{\text{opt}}\) with
the numerical inequalities of III.5.7(b) hold with:
gain \(g = \beta \mu t^2\);
mark \(\mu_s = \mu uP (\beta_F)^2 / 2\) (half of the guaranteed gain is extracted per seed);
\(q = 1 / 2\);
\(\gamma_{\text{root}} = e^{- \mu uP (\beta_F)^2 / 8} + \gamma_{\text{opt}}\).
The packet envelope and optional site-sum estimate in III.5.7(b) remain hypotheses on the actual weights.
(c) For a step-estimator \(\mathcal{T}\) (III.1), there is \(n\) such that for every \(L \geqslant n\) and all sufficiently large \(W\) there is \(\mu > 0\), chosen before \(l, d, u, a, D, z\), with the following property for every choice of \(l, d, u, a, D, z\) satisfying (a). For all sufficiently large \(x\), every funded step with block side \(L\) and coarse side \(\geqslant 100 (1 + B)\) satisfies the point charge of III.5.11 with constant \(\mu\). For every \(\beta \geqslant u \beta_F\), the numerical conclusions of (a) hold with \(g = \beta \mu t^2\). The threshold in \(x\) may depend on the fixed numerical parameters but not on the volume.
Status: theorem. (c) takes the energy grant per seed from III.10. Used in III.10. Lean: Delay.fits, Delay.comparison, Buy.quoted, Grant.order.
Proposition III.5.9 (logarithm by deletion).
(a) Abstract form.
Data. \(J\) finite; \(X : \iota \to\) nonempty subsets of \(J\) (\(\iota\) any set); \(v : \iota \to [0, \infty)\); a weight \(w\) on subsets of \(J\) with \(w \geqslant 1\) and \(w (P \cup P') \leqslant w (P) w (P')\) whenever \(P \cap P' \neq \emptyset\); \(a \geqslant 0\); \(0 \leqslant \delta \leqslant 1 - e^{- a}\).
Site condition. For every \(x \in J\),
Gas and its pieces. For \(u : \iota \to \mathbb{R}\) with \(|u_i | \leqslant v_i\), let \(Z_u (B) = \sum_Q \prod_{i \in Q} u_i\) (\(B \subseteq J\)), the sum over finite families \(Q\) of labels with pairwise disjoint supports contained in \(B\). Let \(\lambda_u\) be the Möbius pieces of \(- \log Z_u\).
Claim.
\(Z_u (B) > 0\) and \(Z_u (B) = \exp \left( - \sum_{A \subseteq B} \lambda_u (A) \right)\) for every \(B\).
\(| \lambda_u (A) | \leqslant \lambda_{- v} (A)\), and \(\sum_{A \ni x} w (A) \hspace{0.17em} \lambda_{- v} (A) \leqslant a\) for every \(x\).
For every \(Q \subseteq J\), \(Z_u (J \setminus Q) = Z_u (J) \hspace{0.17em} e^l\), with \(l = \sum_{A \cap Q \neq \emptyset} \lambda_u (A)\) and \(|l| \leqslant a |Q|\).
(b) At a coarse field.
Data. Let \(\iota\) be countable, with nonempty supports \(X_i\). Let \(E_i (V)\) (\(i \in \iota\)) be functions of the gauge field on \(\Lambda_n\). Each is local in \(X_i\), gauge invariant and Borel, with \(|E_i (V) | \leqslant v_i\) whenever no site of \(X_i\) is flagged. Let \(\tilde{E}_i\) be \(E_i\) set to \(0\) when \(X_i\) contains a flagged site.
Hypotheses. \(M \in \mathbb{N}\), \(A, a \geqslant 0\), \(v_i \geqslant 0\), \(a (2 M + 1)^4 \leqslant 1\), \(\delta \leqslant 1 - e^{- a}\), and \(\sum_{i : \hspace{0.17em} x \in X_i} e^{(A + 1) \text{cov}_M (X_i)} v_i \leqslant \delta\) for every \(x\).
Claim. The gas \(Z_V\) of \(\tilde{E} (V)\) satisfies, for every \(V\) and every \(Q\):
\(Z_V (Q) > 0\) and \(Z_V (Q) = \exp \left( - \sum_{P \subseteq Q} \lambda_P (V) \right)\);
\(| \lambda_P (V) | \leqslant g (P)\), with \(g\) independent of \(V\), and \(\sum_{P \ni x} e^{A \hspace{0.17em} \text{cov}_M (P)} g (P) \leqslant a\);
\(Z_V (\Lambda_n \setminus Q) = Z_V (\Lambda_n) e^l\) with \(l = \sum_{P \cap Q \ne \emptyset} \lambda_P (V)\) and \(|l| \leqslant a |Q|\);
there are gauge-invariant Borel functions \(\lambda_P^{\text{raw}}\), local in the links with both endpoints in \(P\), such that \(\lambda_P (V) =\boldsymbol{1} [P \text{ contains no flagged site}] \lambda_P^{\text{raw}} (V)\). Thus the masked coefficients also use the flag tests on \(P\).
Status: theorem. (b) is used in III.11. Lean: ReLog.catalog, ReLog.omitted, ReLog.swept, ReLog.intake, ReLog.retest, ReLog.retain_mask.
Proposition III.5.10 (trace-log of a near-identity product). Let \(A\) be positive definite and \(C\) symmetric (real matrices), and \(S = AC - 1\) with \(0 \leqslant q < 1\) and \(\sum_j |S_{ij} | \leqslant q\) for every \(i\). Then:
\(\det C > 0\) and \(\det (1 + S) > 0\);
\(\sum_{n \geqslant 1} \frac{(- 1)^{n + 1}}{n} \hspace{0.17em} \text{tr} \hspace{0.17em} S^n\) converges absolutely, and its sum is \(\log \det (1 + S)\).
Status: theorem. Used in III.4. Lean: Round.exact_trace, Round.twin_log.
Proposition III.5.11 (sector decomposition of a block-spin step; gain per seed).
Setting. Fix a step \(S\) with block side \(L\), a step-estimator \(\mathcal{T}\), a tube \(T\) with \(\text{Funds} (S, T)\) (III.1), a coarse coupling \(x\) with threshold \(t\), and constants \(W, \kappa > 0\).
Sector of a packet. Say \((U, k, V)\) satisfies the sector prerequisites of \(A\) when:
each fine cell whose projected vertices meet \(D_{5 B} (A)\) has plaquette chord \(\geqslant t / 4\) if listed in \(A\), and \(\leqslant t / 2\) otherwise;
each coarse bond with an endpoint in \(D_{5 B} (A)\) has predicted wire chord \(\geqslant Wt\) if listed in \(A\), and \(\leqslant 2 Wt\) otherwise;
on \(D_{5 B} (A)\), the sites listed in \(A\) are exactly the sites flagged by \(V\) at the coarse threshold \(t_C = x^{- 1 / 2} e^{(\log \log x)^2}\);
the frame tag has positive frame mass at each site of \(D_{5 B} (A)\).
Point charge. For every tree-gauge forest \(F\), every fine field \(U\) equal to \(1\) on the trunks of \(F\), every frame tag \(k\), every coarse field \(V\) and every packet \(A\) such that \((U, k, V)\) satisfies these prerequisites:
Here \(\mathcal{E}_A\) is the assessed energy of \(A\): the sum of the fine, block-constraint, coarse and cross terms of III.1 with anchors in \(D_{2 B} (A)\), using walks truncated at \(\lfloor B / 100000 \rfloor\).
Claim. Let \(R\) be a lawful representation of the fine density (III.1) adapted to \(T\): its path data and clip are those of the upper stage of \(T\), and its threshold is \(t\). Let its coupling satisfy \(\beta \geqslant 0\). Let \(F\) be any tree-gauge forest and \(D = (D_A)_A\) any allocation: \(D_A (k, U, V) \geqslant 0\) is Borel, \(\sum_A D_A = 1\), and \(D_A \ne 0\) implies the two seed tests above for every fine cell and coarse bond, together with equality of the flag sets on the whole coarse lattice. Positive likelihood then implies the local sector prerequisites. This global support condition is needed to apply the point charge in every lattice orientation. Then for every coarse field \(V\)
where:
\(c_A\) is the \(A\)-th sector output;
\(\pi_{\beta} = Z_{\text{bond}} (\beta)^{- | \text{Bond} (\Lambda_c) |}\), where \(Z_{\text{bond}} (\beta)\) is the single-bond penalty integrated against probability Haar measure;
\(\ell_A (V)\) is the remaining sector integral, including the coarse factor \(\text{sieve} (V)\), with the orientation-averaged assessed energy removed and absolute values on the exposed density;
\(c_A\) and \(\ell_A\) are bounded Borel functions.
Uniformity.
At a committed pair of scales of the run, a single \(\kappa > 0\) serves all volumes, all steps of that pair, and all lawful representations produced by the run whose coupling lies in the admissible band.
For a fixed step-estimator, choose \(L\) sufficiently large, then \(W\) sufficiently large, then \(\kappa > 0\). The point charge holds for all sufficiently large \(x\) and all funded steps of coarse side \(\geqslant 100 (1 + B)\), uniformly in their volumes.
Status: theorem, conditional on \(\text{Funds}\) and the point charge. The point charge is discharged in two ways:
in general, from the Teller's energy grant (III.10);
at committed scales, from the committed-scale energy hypothesis of the run, through notation_ticket (III.1).
Used in III.4 and III.10. Lean: Presented.portionCost (conclusion PortionCost), Consumer.bought, points_grant; definitions Seeded, Alive, PointCharge, earn, Allocation.
Proposition III.5.12 (Combes–Thomas bound). Let \(M\) be a real symmetric matrix on a finite index set, with positions \(a_i \in \Lambda_n\). Assume \(M \geqslant r > 0\), \(\eta \geqslant 0\), and
If \(0 \leqslant \theta \leqslant 1\) and \(\theta C \leqslant r / 2\), then for every \(u \geqslant 0\) the matrix \(M + uI\) is invertible and
The hypotheses pass to principal submatrices. Status: theorem. Lean: Green.inverse, Green.Bottom.restrict, Green.Rows.restrict.
Proposition III.5.13 (windowed inverses).
Setting.
A finite index set, positions \(d_j \in \Lambda_n\), and \(U \geqslant 0\) such that \(\text{card} \{j : d (y, d_j) \leqslant \rho\} \leqslant U (1 + \rho)^4\) for every \(y\) and every \(\rho \in \mathbb{N}\).
\(A, r, C, \eta, \theta\) as in III.5.12, with \(C \geqslant 0\) and \(\eta, \theta > 0\); and \(0 \leqslant \mu < \nu < \theta \eta\).
A radius \(R \in \mathbb{N}\) and windows \(P (j) \ni j\) such that \(d (d_i, d_j) \leqslant R \Rightarrow i \in P (j)\).
Objects.
\(\text{within} (A)_{ij} = ((A|_{P (j)})^{- 1})_{ij}\) for \(i \in P (j)\), and \(0\) otherwise.
\(E = A \hspace{0.17em} \text{within} (A) - 1\) and \(\text{cov} = \tfrac{1}{2} (\text{within} + \text{within}^T)\).
\(\Pi_U (\eta) = U \hspace{0.17em} 4! \hspace{0.17em} (2 / \eta)^4 \hspace{0.17em} e^{\eta / 2} / (1 - e^{- \eta / 2})\), \(Y = \Pi_U (\theta \eta - \nu) / \min (1, r / 2)\), and \(s = e^{- (\nu - \mu) R} (CY + 1)\).
(a) Weighted row and column sums at rate \(\mu\) satisfy:
\(A^{- 1}\) and \(\text{within} (A)\): at most \(Y\);
\(E\): at most \(s\);
\(\text{within} (A) - A^{- 1}\): at most \(Ys\).
At any rate \(0 \leqslant \xi \leqslant \eta\), if \(\text{within} - A^{- 1}\) has row and column bounds \(s' \geqslant 0\) and \(\text{within}\) has bounds \(Y' \geqslant 0\), then \(\text{cov}\), \(\text{cov} - A^{- 1}\) and \(A \hspace{0.17em} \text{cov} - 1\) have bounds \(Y'\), \(s'\) and \(Cs'\), respectively. If also \(C, Y' > 0\) and \(s' \leqslant 1 / (2 C)\), then \(\text{cov} \geqslant 1 / (2 C)\) and \(\text{cov}^{- 1} \geqslant (Y')^{- 1}\).
(b) Fix positive bounds on the weighted row and column sums, spectral gap, rate \(\alpha\), and counting constant. There are \(0 < e \leqslant \alpha\) and \(K, c, c_1, c_2 > 0\), depending only on these data, such that for all sufficiently large integer \(R\) the following holds uniformly for all such matrices and windows with room \(R\). At rate \(e\), the full and windowed inverses of \(tA + (1 - t) I\) have bounds \(K\), and their residual and inverse difference have bounds \(Ke^{- cR} < 1\), for every \(t \in [0, 1]\). At \(t = 1\), the covariance has bound \(K\), its difference from \(A^{- 1}\) and the residual \(A \hspace{0.17em} \text{cov} - I\) have bounds \(Ke^{- cR}\), and \(\text{cov} \geqslant c_1\), \(\text{cov}^{- 1} \geqslant c_2\).
Status: theorem. Lean: Win.Windows.window_price, covariance_price, covariance_floor, Win.Promise.receive.
Proposition III.5.14 (precision matrix of the small-field Gaussian of a step).
Setting. Fix a step-estimator \(\mathcal{T}\) and a patch \(\mathcal{D}\) (block side \(L\), order \(J\)) satisfying the branch conditions for its kernel and gauge group. For a step with patch \(\mathcal{D}\), tube \(T\), coupling \(x\), tree-gauge forest \(F\), packet \(A\) and coarse field \(V\), define \(D\) as follows:
Rows. Use the flat plaquette-curl and upper path-sum rows, with walks truncated at \(\lfloor B / 100000 \rfloor\), and the flat linearized block-constraint rows.
Transport. Transport the input Lie-algebra field by the replica of \(V\) and a fixed path gauge centred at each row, in every lattice orientation.
Restriction. Keep rows whose oriented centres lie outside \(D_{8 B} (A)\), and columns on active fine bonds: bonds outside the trunks of \(F\) whose projected endpoints lie outside \(D_{10 B} (A)\). Set all other entries to zero.
Normalization. Write Lie-algebra coordinates and multiply each row by the inverse square root of the number of lattice orientations.
This is a transported flat linearization. Its equality to the derivative of the nonlinear cut rows at \(V\) is not asserted.
Let \(\text{gram} = D^T D + \text{diag} (\boldsymbol{1}_{\text{inactive}})\), with positions the coarse parents of the bonds.
Claim. There are constants \(r, K, U, \alpha > 0\), with \(\alpha\) half the decay rate of the upper path data, such that the following holds for all large \(x\). For every step with patch \(\mathcal{D}\) and \(\text{Funds}\), coarse side \(\geqslant 100 (1 + B)\), every \(F\), and every \(A, V\) for which the flags of \(A\) are the flagged sites of \(V\):
\(\text{gram}\) is symmetric and \(\text{gram} \geqslant r\);
at most \(U (1 + \rho)^4\) indices lie within distance \(\rho\) of any site;
its weighted row and column sums at rate \(\alpha\) are at most \(K\);
\(D_{ij} = 0\) when the row centre and column position are separated by more than \(\lfloor B / 100000 \rfloor + R_{\text{pred}} (\mathcal{D}) + 2\), where \(R_{\text{pred}}\) is the patch’s predictor range. The range grows with \(B\); the weighted bounds are uniform in \(x\).
Consequently, for large \(x\), III.5.13(b) applies to the following windows: a fine bond belongs to the window of an addressed bond exactly when at least one of its projected endpoints is within distance \(\lfloor B / 4 \rfloor\) of at least one projected endpoint of the addressed bond. Include every color coordinate of each such bond. Their room is \(\lfloor B / 4 \rfloor\), and the error is \(K_{\text{win}} e^{- c_{\text{win}} \lfloor B / 4 \rfloor} < 1\).
Status: theorem, conditional on \(\text{Funds}\) and \(\text{Branches}\). Used in III.4 and III.11 through the window estimates below. Lean: Aff.bedrock, Aff.surplus, Aff.gram_budget, Aff.install_promises; definitions Aff.Dmat, Aff.gram, Promise.Served.
Further outputs.
A calculus of budgets \(\text{SB}\) (Basic.lean): monotonicity, scaling, pull-back, sums over fibres, geometric tails, budget_count_tail. It is used by about 125 modules elsewhere.
Site-sum bounds for bundles of walks on the fine lattice, with a tail in the true length of the walks (Paths.bundle_hit).
A weighted row bound for the tree-gauge response kernel (Response.response_price).
Auxiliary Gaussian estimates, listed in Lemma 4.36.
From other units.
I.1: the torus \(\Lambda_n\); the distance \(d\) (separation); \(|X| \leqslant \text{ball}_M \text{cov}_M (X)\) (Corona.pays); subadditivity of \(\text{cov}_M\) on overlapping unions (Corona.meets_sum).
II.2: the site-sum condition Ac.Hit; connected families Ac.Spark; \(\text{cov}_M\) of a connected union is at most the sum (Ac.cover).
III.1:
\(\text{SB}\);
the step data: patch, kernel, frame choices, the pushforward StepData.push, Branches;
lawful representations Rep.Lawful;
packets, domes, sectors and the assessed energy (Stores, Plot, assessed);
the step-estimator interface Teller and Teller.Funds, with \(\text{Funds}\) supplied for the run by Testing.proceeds;
tubes, tree-gauge forests (Tree.Forest) and flags;
the constants \(B\), \(t\), \(\beta_F\) (radiusB, fiery, betaF);
the point charge at committed scales (notation_ticket).
III.10:
the step-estimator’s energy grant per seed (Next.Fluctuations.grants), used for III.5.8(c) and III.5.11;
the sector integrals and their reduction (Sector.peel, Sector.prefactor);
the linearized block constraint zLin;
the growth of \(\beta_F\) and the uniform predictor bound used in the spectral gap estimate.
III.6: a gauge-fixed Poincaré-type inequality (Work.F01.link_bound). A field vanishing on the trunks is bounded on each fine bond by its plaquette curls in a box of fixed radius around the block, plus its block average. This is the source of the spectral gap \(r\) in III.5.14.
III.2: committed pairs of scales, the construction run and its energy hypothesis entering the uniform sector estimate.
Standard mathematics. The formal proofs establish the estimates directly. References for the methods are:
Mayer expansion, Ursell functions and the Penrose tree partition: R. Fernández and A. Procacci, “Cluster expansion for abstract polymer models. New bounds from an old approach”, Commun. Math. Phys. 274 (2007), 123–140.
Polymer convergence criteria: R. Kotecký and D. Preiss, “Cluster expansion for abstract polymer models”, Commun. Math. Phys. 103 (1986), 491–498; C. Gruber and H. Kunz, “General properties of polymer systems”, Commun. Math. Phys. 22 (1971), 133–161.
Lattice-animal counting by spanning walks; the Schur test for matrix norms; Möbius inversion on subsets.
Exponential conjugation for inverse decay: J.-M. Combes and L. Thomas, “Asymptotic behaviour of eigenfunctions for multiparticle Schrödinger operators”, Commun. Math. Phys. 34 (1973), 251–270.
Large-field sectors and localized propagators belong to Bałaban’s block-spin method. Here they are implemented through the explicit seed allocations and principal inverses stated above.
Imported constants and assumptions.
The constants \(B = \lceil (\log x)^2 \rceil\), \(t = \beta_F^{- 1 / 2} e^{(\log \log \beta_F)^2}\) and \(\beta_F = x + b \log L\) (III.1).
The halo factor \(l\) is arbitrary in the packet estimates; a particular choice belongs to their application.
The block side \(L \geqslant n\) and the constant \(W\), chosen large (III.2).
Assumed hypotheses:
\(\text{Funds}\) and \(\text{Branches}\) (III.1);
lawful representations (III.1);
the point charge (discharged as in III.5.11);
in the auxiliary Gaussian estimates, the coarse-side bound and flag equality, together with the compatible quadratic forms specified in Lemma 4.36.
For a finite gas, inclusion–exclusion writes the partition function as a sum over tuples. Splitting off the connected component of a distinguished occurrence gives the coefficient recurrence \((m + 1) g_{m + 1} = \sum_{k = 0}^m (k + 1) \ell_k g_{m - k}\). For the gas polynomial \(P (t) = \sum_n g_n t^n\) and cluster series \(L (t) = \sum_m \ell_m t^{m + 1}\), this becomes \(P' = L' P\). A geometric bound on the cluster coefficients justifies differentiation on an interval containing \([0, 1]\); then \(Pe^{- L} = 1\). Absolute bounds allow passage to arbitrary label sets through the directed family of finite subsets.
The bounds combine the tree-graph inequality with a sum over trees. A parent support \(Y\) meets labels of total inflated weight at most \(|Y| \gamma \leqslant \text{ball}_B \gamma \hspace{0.17em} \text{cov}_B (Y)\). The reserve exponent \(\delta\) absorbs each parent’s number of children through \(x^k \leqslant k!e^x\). Substochastic kernels sum the labels; the sum of the resulting child factorials over parent maps is at most \(4^m m!\). After cancellation of the occurrence factorial, the ratio per added vertex is \(4 \hspace{0.17em} \text{ball}_B \gamma / \delta\). Subadditivity of the covering number transfers these bounds to cluster unions. The resulting estimates retain site pins, tails, deleted regions and repeated forced polymers.
For seed packets, a spanning-walk count gives an entropy factor \(K_{\text{step}}^2\) per seed. The point charge supplies an energy gain \(\beta \kappa t^2\) per seed. Since \(t = \beta_F^{- 1 / 2} P (\beta_F)\) and \(\beta \geqslant u \beta_F\), the guaranteed gain is \(\kappa uP (\beta_F)^2\). At fixed \(L\), this dominates every cost \(\text{bill}^d P (\beta_F)^z\) with \(z < 2\). Half the gain can be retained as a mark for later estimates. The partition of unity over global seed records then splits the block-spin pushforward while preserving a common field, tag array and tree-gauge slice.
The deletion construction obtains another logarithm from the region recurrence. Its negative-logarithm pieces form a union-convolution series pinned at a chosen site. A weighted \(\ell^1\) bound closes induction on the region and compares signed activities with the gas of negative envelope activities. The final matrix estimates use a transported flat linearization: a gauge-fixed Poincaré inequality and a small-background comparison give a uniform spectral gap, while path bounds give weighted row and column estimates. Exponential conjugation yields inverse decay, and principal window inverses approximate the full inverse with error \(O (e^{- cB})\).
In Lemmas 4.1–4.4 the label set \(\iota\) is finite, the weights are real, and \(R\) is reflexive and symmetric. All vertex sets are finite.
Lemma 4.1 (gas as a sum over tuples). Let \(R\) be reflexive and symmetric, and \(g_n = \frac{1}{n!} \sum_{z \in \iota^n} \text{all} (z) \prod_a w_{z_a}\), where \(\text{all} (z) = \sum_{G \subseteq G_R (z)} (- 1)^{|G|}\). Then \(g_n = \sum_{|s| = n, s \text{compatible}} \prod_{i \in s} w_i\), and \(\sum_n g_n = \Xi (w)\). Sketch:
\(\text{all} (z) =\boldsymbol{1} [G_R (z) = \emptyset]\), by the binomial identity.
Since \(R\) is reflexive, an edgeless tuple has distinct entries, and so orders a compatible set; each set arises \(n!\) times.
Lean: Partition.allMass_eq, Gas.row, Gas.sums.
Lemma 4.2 (splitting off a component). For any relation \(R\), any tuple \(z\) on a finite vertex set \(\alpha\), and \(o \in \alpha\):
Sketch: a subgraph is the same thing as three pieces of data:
the vertex set \(s\) of the component of \(o\);
a connected spanning graph on \(s\);
an arbitrary subgraph on the complement, with no edges between \(s\) and its complement.
The signs multiply. Lean: Partition.partition. (proof read; the bijection lemmas statement only)
Lemma 4.3 (coefficient recurrence). \((m + 1) \hspace{0.17em} g_{m + 1} = \sum_{k = 0}^m (k + 1) \hspace{0.17em} \ell_k \hspace{0.17em} g_{m - k}\). Sketch:
Sum Lemma 4.2 over all tuples on \(\{o\} \sqcup \{1, \ldots, m\}\); the tuple sums over \(s\) and over its complement factor.
There are \(\binom{m}{k}\) sets \(s\) of size \(k + 1\) containing \(o\).
Relabelling invariance of \(\phi^T\) finishes the count.
Lean: Coef.unbased, Coef.discrete, Coef.recurrence.
Lemma 4.4 (III.5.1). Under the geometric coefficient bound of III.5.1, \(\exp (\sum_m \ell_m) = \Xi (w)\). Sketch:
Let \(u = (1 + q^{- 1}) / 2 > 1\). On \((- u, u)\), the series \(L (t) = \sum \ell_m t^{m + 1}\) can be differentiated term by term, because \(| (m + 1) \ell_m t^m | \leqslant C (m + 1) (qu)^m\) is summable.
\(P (t) = \sum g_n t^n\) is a polynomial, since \(g_n = 0\) for \(n > | \iota |\).
Lemma 4.3 and the Cauchy product give \(P' = L' P\). Hence \(H = Pe^{- L}\) has \(H' = 0\) and \(H (0) = g_0 = 1\).
Evaluate at \(t = 1\).
Lean: Gas.exact_polynomial.
Lemma 4.5 (Penrose partition; III.5.2). The absolute connected-graph sum is bounded by the number of spanning trees. Sketch:
To each connected spanning \(G\) attach its breadth-first tree from the distinguished vertex, breaking ties by a fixed enumeration.
For a fixed tree \(p\), the graphs with tree \(p\) form a Boolean interval \([E (p), E (p) \cup F (p)]\). Here \(E (p)\) is the edge set of \(p\), and \(F (p)\) consists of the edges of \(T\) that neither change the breadth-first layers nor offer an earlier parent.
The alternating sum over an interval vanishes unless the interval is a point: flip one free edge.
Each fibre therefore contributes at most \(1\), and only trees of \(T\) contribute.
The bound follows without identifying which trees have nonzero fibres. Lean: Graph.pay; breadth-first construction in Breadth.lean. (proof read; breadth-first construction statement only)
Lemma 4.6 (substochastic sums along a forest). Let \(\alpha, \beta\) be finite, \(p : \alpha \to \alpha \sqcup \{o\}\) an acyclic parent map, \(Y\) a fixed root state, and \(X_j\) the state associated with \(j \in \beta\). If \(K (Y', j) \geqslant 0\) and \(\sum_j K (Y', j) \leqslant 1\) for every state \(Y'\), then \(\sum_{z : \alpha \to \beta} \prod_i K (X_{z_{p (i)}}, z_i) \leqslant 1\), with the first argument interpreted as \(Y\) when \(p (i) = o\). Sketch: induction on the set of summed vertices. Sum out a vertex of maximal rank first: it is a leaf, and its row sum is \(\leqslant 1\). Lean: Trees.substochastic.
Lemma 4.7 (fan-out factorials). \(\sum_{p : [m] \to [m] \sqcup \{o\}} \prod_v (\text{card} \{i : p (i) = v\}) ! = \binom{2 m}{m} m! \leqslant 4^m m!\). Sketch: with \(m + 1\) possible parents fixed, extending a map on \(n\) vertices multiplies the total factorial weight by \(n + m + 1\). Iterating from \(n = 0\) to \(m\) gives \((2 m) ! / m!\), and \(\binom{2 m}{m} \leqslant 4^m\). Lean: Trees.parent_fanout, fanout_bound.
Lemma 4.8 (trees rooted at a support). Assume \(w \geqslant 0\), \(\gamma, \delta > 0\), \(a \in \mathbb{R}\), and \(\text{Hit}_B (w ; a + \delta, \gamma)\). For any support \(Y\) and any finite \(D \subseteq \iota\),
where \(p\) runs over acyclic parent maps on the \(m\) labelled vertices with external root of support \(Y\). Sketch:
Define \(K (Y', i) =\boldsymbol{1} [X_i \cap Y' \neq \emptyset] \hspace{0.17em} w_i e^{(a + \delta) \text{cov}_B (X_i)} / \left( \tau \max (\delta \hspace{0.17em} \text{cov}_B (Y'), 1) \right)\) with \(\tau = \text{ball}_B \gamma / \delta\). The numerator summed over \(i\) is at most \(|Y' | \gamma \leqslant \tau \delta \hspace{0.17em} \text{cov}_B (Y') \leqslant \tau \max (\delta \hspace{0.17em} \text{cov}_B (Y'), 1)\), so \(\sum_i K (Y', i) \leqslant 1\), also for \(Y' = \emptyset\).
Rewrite each weight \(w_i e^{a \hspace{0.17em} \text{cov}}\) as \(\tau \max (\delta \hspace{0.17em} \text{cov} (\text{parent}), 1) \hspace{0.17em} K \hspace{0.17em} e^{- \delta \hspace{0.17em} \text{cov} (X_i)}\).
A parent with \(k\) children contributes \(\max (\delta \hspace{0.17em} \text{cov}, 1)^k \leqslant k! \hspace{0.17em} e^{\delta \hspace{0.17em} \text{cov}}\). The exponential at each internal vertex cancels its own \(e^{- \delta \hspace{0.17em} \text{cov}}\), leaving \(e^{\delta \hspace{0.17em} \text{cov}_B (Y)}\) at the root.
Conclude with Lemmas 4.6 and 4.7.
Lean: Trees.trees_afford, Trees.forest_at.
Lemma 4.9 (rooted and marked cluster sums; III.5.3(c)). Under III.5.3’s hypotheses, the marked sum at order \(m\) is at most \(bq^m\). Sketch:
\(\frac{1}{(m + 1) !} = \frac{1}{m + 1} \cdot \frac{1}{m!}\): write the tuple sum as the average over the choice of root vertex. A transposition moves the root to position \(0\), and \(\phi^T\) is invariant under relabelling.
At root \(0\), Lemma 4.5 bounds \(| \phi^T (z) |\) by the trees of \(G_R (z)\). Since \(R\) is contained in the overlap relation, these are trees of supports rooted at \(X_{z_0}\) in the sense of Lemma 4.8.
The root label is summed against its own weight \(w_i e^{(a + \delta) \text{cov} (X_i)}\), which absorbs the factor \(e^{\delta \hspace{0.17em} \text{cov}}\). This gives \(bq^m\).
Lean: Ursell.cluster_at.
Lemma 4.10 (pinned bounds; III.5.3(a),(b)). Under III.5.3’s hypotheses, both pinned tails are at most \(\gamma q^m / (1 - q)\). Sketch:
If \(\phi^T_R (z) \neq 0\), then \(G_R (z)\) is connected, so \(\text{cov}_B (X (z)) \leqslant \sum_a \text{cov}_B (X_{z_a})\).
Apply Lemma 4.9 with \(Q = \{i : x \in X_i \}\), where \(b = \gamma\) by the site condition at \(x\), and sum the geometric tail.
The set version (b) follows in the same way from a count of distinct-label forests.
Lean: Ursell.connected_pay, Ursell.log_hit_tail, Trees.touch_tail. (proof read; distinct-label forest count statement only)
Lemma 4.11 (from finite to arbitrary label sets; III.5.4). Under III.5.4’s hypotheses, the gas and cluster series converge absolutely, and \(\Xi = \exp (\text{cluster series}) > 0\). Sketch:
Every support is nonempty, so \(\sum_i |w_i | e^{(a + \delta) \text{cov}} \leqslant n^4 \gamma\). Hence \(\sum_s \prod_{i \in s} |w_i | \leqslant e^{n^4 \gamma}\), and the cluster series has absolute sum \(\leqslant n^4 \gamma / (1 - q)\).
On a finite \(D \subseteq \iota\) the hypotheses restrict, and III.5.1 applies with \(C = n^4 \gamma\).
Let the finite subsets \(D \subseteq \iota\) increase by inclusion and use dominated convergence for both series; the dominating budgets do not depend on \(D\).
Lean: Gas.finite_log, Gas.price_log, Product.everyone_ne, Product.set_products.
Lemma 4.12 (III.5.5). Under III.5.5’s hypotheses, deleting polymers meeting \(Y\) multiplies the gas by \(e^{- p}\), with \(|p| \leqslant |Y| \gamma / (1 - q)\) and the stated cluster tails. Sketch:
The cluster terms of \(w^Y\) are those of \(w\) minus the terms whose union meets \(Y\).
The removed terms are bounded by Lemma 4.10 with \(m = 0\), summed over \(x \in Y\).
Apply III.5.4 to both gases; \(|w^Y | \leqslant |w|\), so \(w^Y\) satisfies the same hypotheses.
Lean: Gas.restores_log, Gas.missing_tail.
Lemma 4.13 (multiplicities on a fixed set). For a finite set \(s\), if \(0 \leqslant \sigma < 1\) and \(0 \leqslant w_i \leqslant \sigma\) for \(i \in s\), then \(\sum_{(n_i) \in \mathbb{N}_{\geqslant 1}^s} \prod_{i \in s} w_i^{n_i} / n_i ! \leqslant \prod_{i \in s} (w_i / (1 - \sigma))\). Sketch: bound \(w_i^{n_i} / n_i !\) by \(w_i \sigma^{n_i - 1}\) and multiply the geometric sums. Lean: Multiple.buy_at.
Lemma 4.14 (multiset sums). Let \(X_i\) be nonempty, \(w_i \geqslant 0\), \(a \geqslant 0\), \(t \geqslant 1\), \(0 \leqslant \sigma < 1\), and \(e_i \in \mathbb{R}\). Suppose \(\text{Ready} \left( t \hspace{0.17em} w \hspace{0.17em} e^e / (1 - \sigma) ; a, \gamma, \delta, q \right)\) and \(\gamma \leqslant \sigma\). Then the sum over multisets with connected support through \(x\), satisfying a predicate \(Q\) on which \(\sum n_i e_i \geqslant v\), with total multiplicity \(\geqslant m\), of \(e^{a \hspace{0.17em} \text{cov}_B (\cup)} \prod w_i^{n_i} / n_i !\), is at most \(t^{- m} e^{- v} \gamma / (1 - q)\). Sketch:
Each weight \(tw_i e^{e_i} e^{a \hspace{0.17em} \text{cov}_B (X_i)}\) is at most \(\sigma\), by the site condition at a point of its nonempty support.
Connectedness bounds the union cover by the sum of the individual covers; multiplicities are at least one. Lemma 4.13 then sums the multiplicities at cost \(1 / (1 - \sigma)\) per label, and III.5.3(b) sums the connected sets.
The factor \(t^{\sum n} e^{\sum ne} \geqslant t^m e^v\) is extracted on the selected multisets.
Lean: Multiple.estimates.
Lemma 4.15 (III.5.6). Under III.5.6’s hypotheses, the pinned multiset sum containing a forced label is at most \(t^{- m} e^{- \mu} \eta / (1 - q')\). Sketch:
The first-level clusters satisfy a site condition with constant \(\gamma / (1 - q)\) at exponent \(A + \delta'\), by III.5.3(a) with \(m = 0\); the forced polymers satisfy one with constant \(c\).
Join the two alphabets, scaling the weights by \(te^{\mu} / (1 - \sigma)\) and \(t / (1 - \sigma)\) respectively; the resulting constant is at most \(\eta\).
Apply Lemma 4.14 with \(e = \mu\) on forced labels; the presence of a forced label gives \(v = \mu\).
Lean: OmitClusters.restore, Labels.together.
Lemma 4.16 (animals and covers). In the setting of III.5.7:
For every \(k \in \mathbb{N}\), the number of connected packets with \(k\) seeds whose dome contains \(x\) is at most \(K_{\text{hit}} (K_{\text{step}}^2)^k\).
\(\text{cov}_B (D (A)) \leqslant |A| \hspace{0.17em} (2 l + 3)^4\) for connected \(A\).
Sketch:
A connected packet is the vertex set of a walk of length \(2 (k - 1)\). The walk runs in the graph of seeds whose skins can meet, which has degree \(\leqslant K_{\text{step}}\). It starts at one of at most \(K_{\text{hit}}\) seeds whose skin contains \(x\).
For \(B \geqslant 1\), a skin lies in a ball of radius \(lB + 1 \leqslant (l + 1) B\), covered by \((2 l + 3)^4\) balls of radius \(B\). For \(B = 0\), its connected location set lies in the radius-\(1\) ball and costs at most \(3^4 \leqslant (2 l + 3)^4\) singleton balls. A connected union costs at most the sum.
Lean: Packets.stores_count, Animals.animals, Packets.cover_support. (statement only for the animal count; covering proof read)
Lemma 4.17 (III.5.7(a)). Under III.5.7(a)’s hypotheses, the packet tail is at most \(K_{\text{hit}} Cq^{\max (1, m)} / (1 - q)\). Sketch:
On a connected packet, \(p_A e^{a \hspace{0.17em} \text{cov}} \leqslant C (e^{s - g + a (2 l + 3)^4})^{|A|}\).
With the count of Lemma 4.16, sum the geometric series (budget_count_tail).
Connected packets have at least one seed; hence \(\max (1, m)\).
Lean: Packets.packet_income, Packets.cover_income.
Lemma 4.18 (III.5.7(b)). Under III.5.7(b)’s hypotheses, the mixed-word sum has the bound stated there. Sketch:
Ready. Inflate the weights to \(p_A e^{2 \varepsilon e_s |A| + \mu_s |A|}\) and \(w_i e^{2 \varepsilon \lambda_i}\). The joint alphabet then satisfies Ready at exponent \(a + 2 \varepsilon\): packets by Lemma 4.17, optional labels by hypothesis, and the constants add up to \(\gamma_{\text{root}}\).
Connected sets. III.5.3(b) sums the connected sets.
Extractions.
\((1 + \text{ch})^k \leqslant \varepsilon^{- k} k! \hspace{0.17em} e^{\varepsilon} e^{\varepsilon \hspace{0.17em} \text{ch}}\);
one of the two factors \(e^{\varepsilon \hspace{0.17em} \text{ch}}\) of the inflation gives \(e^{\varepsilon T}\);
the marks give \(e^{\mu_s r}\).
Lean: Mixed.ready_vertices, Mixed.mixed_hit, Reserves.sets, Reserves.reserve_tail.
Lemma 4.19 (III.5.8). For fixed numerical parameters satisfying III.5.8, all its packet and tree conditions hold for sufficiently large \(x\), uniformly in volume. Sketch:
Domination. \(D \hspace{0.17em} \text{bill}^d \leqslant P (\beta_F)^e\) eventually, for every \(e > 0\), since \((\log \beta)^{2 d} = e^{2 d \log \log \beta}\) is eventually below \(e^{e (\log \log \beta)^2}\). Also \(K_{\text{hit}}, K_{\text{step}}^2 \leqslant e^{M / 16}\), because \(K_{\text{step}} \leqslant \text{const} \cdot \text{bill}^4\).
Ratio. \(\rho \leqslant e^{M / 16} e^{M / 16 - M + M / 16} \leqslant e^{- M / 2}\).
Gain. \(\beta \geqslant u \beta_F\) gives \(\beta \mu t^2 \geqslant \mu uP (\beta_F)^2\), because \(\beta_F t^2 = P (\beta_F)^2\).
Assembly. For (b), subtract the mark \(M / 2\) from the gain and apply (a) with \(\mu / 2\); the remaining gain is at least \(M / 2\). For (c), use the step-estimator’s grant before applying the numerical estimates.
Lean: Delay.monitor_pays, Delay.step_delay, Delay.fits, Buy.quoted, Grant.order. (proof read; the grant of III.10 statement only)
Lemma 4.20 (deletion identity). For the gas \(Z_u\) of III.5.9(a) and \(x \in B\):
and \(Z_u (\emptyset) = 1\). Sketch:
At most one member of a disjoint family contains \(x\); removing it removes its support.
The identity holds exactly for finite truncations of \(\iota\) and passes to the absolutely convergent limit.
Lean: ReLog.stock_deletion, ReLog.shelf.
Lemma 4.21 (the incidence algebra). Let \(J\) be finite and let \(w\) satisfy the weight assumptions of III.5.9(a). For set functions on \(J\) define:
\((f \ast g) (T) = \sum_{P \cup Q = T} f (P) g (Q)\);
\(\hat{f} (B) = \sum_{A \subseteq B} f (A)\), so that \(\widehat{f \ast g} = \hat{f} \hat{g}\);
\(\|f\|_w = \sum_A w (A) |f (A) |\).
Then:
\(\| \cdot \|_w\) is submultiplicative on functions supported on sets containing a common point \(x\).
For such \(f\) with \(\|f\|_w < 1\), \(\lambda (f) = \sum_{n \geqslant 0} f^{\ast (n + 1)} / (n + 1)\) satisfies \(\widehat{\lambda (f)} = - \log (1 - \hat{f})\) and \(\| \lambda (f)\|_w \leqslant - \log (1 -\|f\|_w)\).
If \(f\) and \(F\) are supported on sets containing \(x\), \(|f| \leqslant F\) pointwise and \(\|F\|_w < 1\), then \(| \lambda (f) | \leqslant \lambda (F)\).
Sketch: cumulative sums turn union convolution into multiplication. Two nonzero factors pinned at \(x\) have intersecting supports, so the weight inequality gives submultiplicativity. Apply it to convolution powers and sum \(\sum_{n \geqslant 1} q^n / n = - \log (1 - q)\) for \(q < 1\); termwise absolute domination gives the last assertion.
Lean: ReLog.crop_glue, ReLog.lamp_shed, ReLog.lamp_cropped, ReLog.lamp_mono.
Lemma 4.22 (the induction). Let \(J\) be finite, \(w\) satisfy III.5.9(a)’s weight assumptions, and \(a \geqslant 0\). Let \(U, z\) be functions of supports with \(U \geqslant 0\), \(|z| \leqslant U\), and \(\sum_{P \ni x} w (P) U (P) e^{a |P|} \leqslant 1 - e^{- a}\) for every \(x\). Suppose \(Z, Y\) satisfy the deletion identities with activities \(- U, - z\), respectively, and \(Z (\emptyset) = Y (\emptyset) = 1\). Then for every \(B\):
\(Z (B) > 0\) and \(Y (B) > 0\);
\(| \lambda_Y (B) | \leqslant \lambda_Z (B)\);
\(\sum_{A \subseteq B, A \ni x} w (A) \lambda_Z (A) \leqslant a\) for \(x \in B\).
Sketch: strong induction on \(B\). The empty region is immediate. For nonempty \(B\), pick \(x \in B\).
Define \(f_x\) on subsets of \(B\), extended by zero on other supports, by Möbius inversion from
The conditions \(x \in P \subseteq A\) make \(f_x\) supported on sets containing \(x\). The deletion identity gives \(Z (A) = Z (A \setminus \{x\}) (1 - \hat{f}_x (A))\).
By submultiplicativity of \(w\) and the induction hypothesis at the points of \(P\), \(\|f_x \|_w \leqslant \sum_{P \ni x} w (P) U (P) e^{a |P|} \leqslant 1 - e^{- a}\). Hence \(Z (B) > 0\).
The pieces of \(- \log Z\) on sets containing \(x\) are \(\lambda (f_x)\), with \(\| \lambda (f_x)\|_w \leqslant - \log e^{- a} = a\).
The same computation for \(Y\), with \(|f^Y_x | \leqslant f_x\), gives the domination.
Lean: ReLog.leash, ReLog.matures, ReLog.swept, ReLog.catalog.
Lemma 4.23 (deleting sites). Under III.5.9(a)’s hypotheses, with \(Z = Z_u\) and \(\lambda = \lambda_u\), \(- \log Z (J \setminus Q) = \sum_{A \cap Q = \emptyset} \lambda (A)\), hence \(l = \sum_{A \cap Q \neq \emptyset} \lambda (A)\), and \(|l| \leqslant \sum_{x \in Q} \sum_{A \ni x} w (A) \lambda_{- v} (A) \leqslant a |Q|\) because \(w \geqslant 1\). Lean: ReLog.omitted.
Lemma 4.24 (at a coarse field; III.5.9(b)). Under III.5.9(b)’s hypotheses, its masked logarithm coefficients satisfy the common envelope and deletion-ratio bounds. Sketch:
\(e^{A \hspace{0.17em} \text{cov}_M}\) is an admissible weight, since \(\text{cov}_M\) is subadditive on overlapping unions.
\(a (2 M + 1)^4 \leqslant 1\) gives \(a |X| \leqslant \text{cov}_M (X)\), which converts the condition into III.5.9(a).
Masking. If \(P\) contains a flagged site \(y\), the masked gas does not see \(y\), so its Möbius piece at \(P\) vanishes.
The raw pieces depend only on labels supported in \(P\), so they are local in \(P\) and gauge invariant. Countability makes the gas sums Borel; finite Möbius inversion preserves this. Multiplying by the healthy-region indicator gives the masked coefficients, including the external flag tests.
Lean: ReLog.intake, ReLog.native_rate, ReLog.permits, ReLog.retain_mask, ReLog.retest_meas.
Lemma 4.25 (trace-log; III.5.10). Under III.5.10’s hypotheses, the trace-log series converges absolutely to \(\log \det (I + S)\), and both determinants are positive. Sketch:
\(AC = A^{1 / 2} (A^{1 / 2} CA^{1 / 2}) A^{- 1 / 2}\) is similar to a symmetric matrix.
Its eigenvalues \(\lambda\) are those of \(1 + S\), and \(| \lambda - 1| \leqslant \|S\|_{\infty \to \infty} \leqslant q\).
Hence \(\log \det (1 + S) = \sum \log \lambda = \sum_n (- 1)^{n + 1} \text{tr} \hspace{0.17em} S^n / n\), and \(\det C = \det (1 + S) / \det A > 0\).
Lean: Round.twins, Round.twin_log.
Lemma 4.26 (allocations and columns). An allocation is the partition of unity with global seed tests specified in III.5.11. It satisfies \(0 \leqslant D_A \leqslant 1\); where the likelihood is positive, \(D_A \ne 0\) implies the local sector prerequisites.
The hard allocation, the indicator of the actual booking, is one example.
A smooth decreasing cutoff \(\chi : \mathbb{R} \to [0, 1]\), equal to \(1\) on \((- \infty, 1]\) and \(0\) on \([4, \infty)\), gives another allocation when \(W, t > 0\).
For every coarse \(V\) and \(\beta \in \mathbb{R}\), if the step satisfies its branch conditions and \(g\) is integrable and gauge invariant, then \(\sum_A \text{out}_A (V) = \text{push}_{\beta} g (V)\).
Sketch: sum the allocation before integrating, then apply the tree-gauge integration identity for \(g\) and the gauge-covariant block kernel.
Lean: Allocation, Allocation.hard, Chi.canonical, push_columns.
Lemma 4.27 (cut-out identity). For a representation adapted to the tube as in III.5.11, let \(\Delta_{\text{cut}} (U, V)\) be the orientation-averaged fine cut-row action minus its coarse counterpart. Then
In the input action, split the cut rows of the upper stage, averaged over the lattice symmetries, into two parts:
their coarse counterpart at \(V\), which gives the factor \(\text{sieve} (V)\), taken out of the integral;
the difference \(\Delta_{\text{cut}} (U, V)\), which is kept with the block kernel.
The rest of the action and the polymer gas form \(\text{exposed} (U)\).
Lean: Presented.expel, Presented.use_gap.
Lemma 4.28 (energy factorization). For a step satisfying its branch conditions,
where bars denote averages over the lattice symmetries, applied to the rotated packet. On the tree-gauge slice, the global seed tests and positive likelihood imply the local prerequisites in every orientation. The point charge therefore gives \(\bar{\mathcal{E}}_A \geqslant \kappa t^2 |A|\) wherever \(D_A \ne 0\) and the likelihood is positive; zero-likelihood terms vanish. Lean: Oriented.enters_eq, Oriented.mean_pays.
Lemma 4.29 (III.5.11). Under III.5.11’s funding, adaptation and point-charge assumptions, the sector outputs sum to the pushforward and satisfy its per-seed bound. Sketch:
Express each signed sector output using Lemma 4.27; its sum is the pushforward by Lemma 4.26.
Bound its absolute value by the integral of the absolute exposed density. Apply Lemma 4.28 with \(\beta \geqslant 0\) to extract \(\pi_{\beta} e^{- \beta \kappa t^2 |A|}\).
Bounded Borel factors give bounded Borel sector integrals at each fixed volume.
Lean: Oriented.turn_charge, Presented.portionCost.
Lemma 4.30 (one \(\kappa\)). The uniform point-charge conclusions of III.5.11 hold. Sketch: the general conclusion uses the step-estimator’s energy grant (III.10). At a committed pair of scales, use the run’s energy hypothesis (III.1); its compatibility conditions identify the representation’s paths, clip and threshold with those of the tube. The constant is chosen before the volume, step and representation.
Lean: points_grant, Consumer.bought, charged_at. (proof read; the inputs from III.1 and III.10 statement only)
Lemma 4.31 (III.5.12). Under III.5.12’s hypotheses, the shifted inverse has the stated pointwise exponential bound. Sketch: fix \(j\) and put \(s_k = e^{\theta \eta d (a_k, a_j)}\) and \(w = s \cdot (M + uI)^{- 1} e_j\).
Then \((M + uI + F) w = e_j\), with \(F_{kl} = M_{kl} (s_k / s_l - 1)\).
By convexity of \(\exp\), \(|s_k / s_l - 1| \leqslant \theta e^{\eta d (a_k, a_l)}\). The Schur test then gives \(| \langle w, Fw \rangle | \leqslant \theta C |w|^2 \leqslant \frac{r}{2} |w|^2\).
Hence \(\min (1, r / 2) (1 + u) |w|^2 \leqslant w_j \leqslant |w|\), and \(|w_i | \leqslant |w|\) gives the claim.
Lean: Green.inverse, Green.diminishing, Green.schur.
Lemma 4.32 (III.5.13). Under III.5.13’s hypotheses, the window residual, inverse difference and covariance satisfy its weighted bounds. Sketch:
\(E_{ij} = 0\) for \(i \in P (j)\), because column \(j\) of \(\text{within} (A)\) solves the system on the window. So \(E\) lives at distances \(> R\).
Pointwise decay at rate \(\theta \eta\), from III.5.12 applied to \(A\) and to its principal submatrices, becomes row-sum decay at rate \(\nu\) by the counting bound. This costs \(\Pi_U (\theta \eta - \nu)\).
The far entries of \(E\) at the slower rate \(\mu\) pay \(e^{- (\nu - \mu) R}\).
Finally \(\text{within} - A^{- 1} = A^{- 1} E\), and the weighted Schur classes are closed under products.
Part (b) applies (a) uniformly along \(tA + (1 - t)\).
Lean: Windows.stitched, Windows.inverse_price, Windows.window_price, Win.planePrice.
Lemma 4.33 (spectral gap). Fix \(\mathcal{T}, \mathcal{D}\) as in III.5.14. There is \(0 < r \leqslant 1\) such that, for all sufficiently large \(x\), all funded steps with this patch, coarse side \(\geqslant 100 (1 + B)\) and the stated flag equality satisfy \(r |v|^2 \leqslant |Dv|^2\) for every \(v\) supported on active coordinates. Sketch:
Take an active bond. Its projected endpoints lie outside \(D_{10 B} (A)\), so in an axial gauge the background \(V\) is within \(\delta (x) \to 0\) of \(1\) near the bond.
The Poincaré-type inequality of III.6 bounds the bond variable by flat curls and block averages near the block. These are the flat rows.
In this gauge, the transported flat rows differ from the untransported flat rows by \(O (\delta) |v|_{\text{loc}}\).
Square and sum with bounded overlap: \(|v|^2 \leqslant \mathcal{V}|Dv|^2 + o (1) |v|^2\), which gives \(r = (2 (1 +\mathcal{V}))^{- 1}\).
Then \(\langle v, \text{gram} \hspace{0.17em} v \rangle = |Dv|^2 + |v_{\text{inactive}} |^2 \geqslant r |v|^2\).
Lean: Aff.inflow, Aff.surplus, Aff.decomposes. (proof read; per-bond comparison and III.6 inequality statement only)
Lemma 4.34 (decay). Fix \(\mathcal{T}, \mathcal{D}\) as in III.5.14. There is \(K > 0\), uniform over all funded steps with that patch, all \(V, A, F\) and every real \(x\), such that \(D\) and its unmasked version have weighted row and column bounds \(K\) at rate \(\alpha\) equal to half the upper path-decay rate, and \(\text{gram}\) has bounds \(K^2 + 1\). Moreover \(\text{gram}_{ij} = 0\) at distances greater than \(2 (\lfloor B / 100000 \rfloor + R_{\text{pred}} (\mathcal{D}) + 2)\). Sketch:
Pointwise exponential decay of the row coefficients comes from the path-sum bounds of the upper stage and from the block-constraint rows.
The counting bound converts it into row sums.
Products and sums give \(\text{gram}\).
Lean: Aff.gram_budget, Aff.stencil, Aff.received.
Lemma 4.35 (III.5.14). The uniform matrix and window conclusions of III.5.14 hold. Sketch: combine Lemmas 4.33 and 4.34 with the index-counting bound, then apply III.5.13(b). The windows have room \(\lfloor B / 4 \rfloor \to \infty\) as \(x \to \infty\).
Lean: Aff.bedrock, Aff.install_promises.
Lemma 4.36 (auxiliary Gaussian estimates). Fix the step-estimator and patch of III.5.14. The auxiliary Gaussian (Mint) uses the completed precision \(G = D^T D + I_{\text{inactive}}\), the affine constant \(h\) (including the clipped frozen coordinates), source \(D^T h\), and selected mean \(\mu = - \text{within} (G)^T D^T h\).
For the field estimates below, assume a funded step with this patch, coarse side \(\geqslant 100 (1 + B)\), and the flag equality of III.5.14. Write \(H_{n, p} = B^p t^n\). A reserve \(C_{\text{res}} > 0\) supplies uniform bounds on the upper and lower row sums and the coarse predicted fields. A compatible pair of quadratic forms means upper and lower forms attached to the tube, satisfying its diagonal and renormalization identities, the two quadratic bounds and orientation covariance. The shift and plateau estimates assume such a pair and the window bounds of III.5.13(b). Each large-\(x\) threshold is uniform in the remaining volume and field data.
Aff.escape: With a compatible pair, for every coarse Lie-algebra field \(b\) and every tree-gauge forest, the stationary linear extension solves the untruncated flat fine-row and block-constraint equations with the predicted coarse linear rows as right-hand sides, in every orientation. This is an identity for those linear operators; it does not assert that \(D \mu + h = 0\).
Aff.velocity, Aff.shifted: Assume a reserve. Let \(o \not\in D_{20 B} (A)\) and let a gauge transform of \(V\) agree, on bonds anchored within \(6 B\) of \(o\), with \(\exp b\), where \(\|b\|_{\infty} \leqslant H_{1, 7}\). Let \(v\) be the inverse-gauge transport of the stationary extension minus the replica extension, restricted to active coordinates. Then \(| \mu_i - v_i | \leqslant H_{2, 27}\) whenever the column position is within \(2 B\) of \(o\). At every row centred outside \(D_{20 B} (A)\), \(| (D \mu + h)_i | \leqslant H_{2, 29}\).
Jewels.night: Assume a reserve \(C_{\text{res}}\), \(W > 0\), and the two numerical margins \(C_{\text{res}} / L^2 \leqslant s_0\) and \(C_{\text{res}} + s_0 \leqslant (W / 4) \sqrt{p_0}\), where \(s_0 = \sqrt{c_{\text{direct}}} / 64\), \(c_{\text{direct}}\) is the upper direct quadratic coefficient, and \(p_0 > 0\) is the kernel penalty. There is \(\varepsilon > 0\) such that, for all sufficiently large \(x\), the following holds when the frozen cutoff scale equals \(Bt\). Reconstruct the fine field by inserting \(\exp (\mu + z)\) relative to the replica of \(V\) on active bonds and keeping the fixed tree-gauge slice field on the other bonds. For any fluctuation \(z\), a fine cell whose projected anchor lies outside \(D_{20 B} (A)\) has plaquette chord \(< t / 4\) if \(|z_i | \leqslant \varepsilon t\) on coordinates anchored within \(B\) of that anchor. The same test at the anchor of a coarse bond outside \(D_{20 B} (A)\) gives wire chord \(< Wt\) and makes both likelihood cutoffs equal to \(1\).
Aff.tap_open: Fix an order \(p \in \mathbb{N}\) and \(\delta > 0\). For all sufficiently large \(x\), if every fine bond anchored within \(B / 2\) of a given coarse-bond anchor has chord at most \(H_{1, p}\), then the frame and connector use their minimizing branches and the predicted coarse-bond chord is \(< \delta\), for every tag. The uniqueness of those minimizers is part of the branch hypotheses.
Aff.perturb: There is \(K > 0\), depending only on the step-estimator and patch, such that for every truncation radius \(R\) and every unmasked transported row \(L_{R, V}\), with \(s_R = R + R_{\text{pred}} (\mathcal{D}) + 2\), \(\|(L_{R, V} - L_{R, 1}) u\| \leqslant K (4 s_R + 1) \delta C_u\) if \(C_u, \delta \geqslant 0\), \(\|u_j \| \leqslant C_u\) on fine bonds anchored within \(s_R + 1\) of that row, and all coarse-bond chords anchored within \(4 s_R + 1\) are at most \(\delta\). This compares with the flat background; it is not a Lipschitz estimate between arbitrary backgrounds.
Aff.arc_bill: Assume a reserve and fix \(p \geqslant 7\). For all sufficiently large \(x\), take a row centred outside \(D_{2 B} (A)\) and a fine Lie-algebra field \(u\) bounded by \(H_{1, p}\) on bonds anchored within \(B\). If \(U\) agrees there with \(\exp (u) \hspace{0.17em} \text{replica} (V)\), its nonlinear row value \(F (U, V)\) satisfies \(\|F (U, V)\| \leqslant H_{1, p + 4}\) and \(\|F (U, V) - F (\text{replica} (V), V) - L_V u\| \leqslant H_{2, 2 p + 6}\), where \(L_V\) is the transported flat row at the prescribed truncation. This is the error estimate used in place of identifying \(L_V\) with a derivative at \(V\).
Aff.reserves: For every step-estimator there exists a reserve \(C_{\text{res}}\). It is chosen before \(L\); for each fixed \(L\), its row and coarse-field bounds hold for all sufficiently large \(x\) and all funded steps of coarse side \(\geqslant 100 (1 + B)\).
The tree-graph argument is standard. The estimates must also retain arbitrary label sets, repeated occurrences, site pins and covering weights at the growing scale \(B = \lceil (\log x)^2 \rceil\). The entropy factor \(4 (2 B + 1)^4\) is sufficient because it grows only polynomially in \(\log x\).
The gain scale is \(P (\beta_F)^2 = e^{2 (\log \log \beta_F)^2}\). It absorbs \(\text{bill}^d P (\beta_F)^z\) for fixed \(d\) and \(z < 2\), but grows more slowly than every positive power of \(\beta_F\). The deletion route gives support-indexed logarithm coefficients and a common field-independent envelope; the graph route also gives occurrence tails and multiset bounds. Both give deletion ratios uniform in volume, although the full logarithm is extensive.
The sector estimate needs global seed records to preserve the local point-charge prerequisites under lattice orientations. Its remainder \(\ell_A\) has no volume-uniform bound here. For the Gaussian estimates, the substantive inputs are the gauge-fixed Poincaré inequality and control of the transported flat rows away from the flagged region. Finite range grows with \(B\); uniform exponential row bounds provide the useful control.
| Written proof | Lean | Meaning |
| \(\text{SB} (f ; C)\) | SumBudget | \(f \geqslant 0\) summable, \(\sum f \leqslant C\) |
| \(\text{Hit}_B (w ; a, \gamma)\) | Ac.Hit | site-sum condition |
| \(\text{cov}_B\), \(\text{ball}_B\) | coverNumber, ballCost | connected covering number; \((2 B + 1)^4\) |
| \(\text{Ready} (w ; a, \gamma, \delta, q)\) | Trees.Ready, degreePrice | smallness of the cluster expansion |
| \(R \subseteq\) overlap | Ursell.Tests, blocked | incompatibility implies overlap; the overlap relation |
| \(G_R (z)\), \(\phi^T_R (z)\) | Ursell.pot, Ursell.coefficient | incompatibility graph of a tuple; Ursell coefficient |
| \(\Xi (w)\); cluster term; log series | Gas.total, Gas.term, Gas.logSum | gas, its cluster terms and their sum |
| \(g_n\), \(\ell_m\) | Coef.gasRow, Coef.logRow | coefficients of \(\Xi\) and of the log series |
| \(w^Y\), \(p\) | Gas.cut, Gas.missing | weights off \(Y\); clusters meeting \(Y\) |
| multisets, \(\prod w^n / n!\) | Multiple.Reps, manyW | labels with multiplicities |
| packet \(A\), number of seeds, \(D_r (A)\) | Stores, shares, dome | finite set of typed seeds; radius-\(r\) neighbourhood of their locations |
| connected packet, \(K_{\text{hit}}\), \(K_{\text{step}}\), \(\tau_L\) | Packets.connected, hitCost, stepCost, typesAt | animal-counting constants |
| \(\rho\), packet tail | ratioCost, fit, totalCharge | per-seed ratio; tail of packet sums |
| hypotheses of III.5.7(b) | Mixed.Costs.Buy, Mixed.Accounts | numerical list; weight bounds |
| \(B\), \(\beta_F\), \(t\), \(P\) | radiusB, betaF, fiery, Vac.Numbers.P | \(\lceil (\log x)^2 \rceil\); \(x + b \log L\); \(\beta_F^{- 1 / 2} P (\beta_F)\); \(e^{(\log \log s)^2}\) |
| \(M\), gain \(\beta \mu t^2\) | hatWin, Grant.credit | guaranteed and actual gain per seed |
| \(Z (B)\), \(\lambda (A)\) | fumes, locker, pieces | disjoint-family gas; Möbius pieces of \(- \log Z\) |
| \(f \ast g\), \(\hat{f}\), \(\|f\|_w\), \(\lambda (f)\) | glue, shed, crop, lamp | union convolution, cumulative sum, weighted norm, log series |
| hypotheses of III.5.9 | Rate, Deletion, Receipt, Sufficient, Fund, Tags | weight, deletion identity, site conditions, coefficient properties |
| allocation \(D\); point charge | Allocation, PointCharge | partition of unity over packets; \(\mathcal{E}_A \geqslant \kappa t^2\) per seed |
| \(\mathcal{E}_A\); \(c_A\); \(\ell_A\); \(\pi_{\beta} e^{- \beta \kappa t^2 k}\) | assessed, cutEntry, leftOver, earn | sector energy; sector output; remainder; gain |
| weighted Schur classes; \(M \geqslant r\) | Green.Rows, Green.Both, Green.Bottom | row/column sums at rate \(\eta\); lower bound |
| windows, \(\text{within}\), \(E\), \(\text{cov}\), \(\Pi_U\) | Win.Windows, within, error, cov, planePrice | localized inverse and covariance |
| uniform matrix data; window constants | Win.Promise, Turns, Turns.Takes | (mass, bottom, rate, counting); bounds of III.5.13(b) |
| \(D\), \(\text{gram}\) | Aff.Dmat, Aff.gram | transported flat row matrix; completed precision matrix |
| transported flat derivative | Aff.derivative | row operator used to define \(D\) |
| auxiliary Gaussian assumptions | Aff.Reserve, Aff.Alongside, Aff.cert | reserve bounds; compatible quadratic forms; volume and flag tests |
| \(H_{n, p}\); comparison field \(v\) | Rates.allowance, Aff.unit | \(B^p t^n\); active stationary extension minus replica extension, transported back by the local gauge |
Imported analytic inputs. The step-estimator’s energy grant in III.10 and the Poincaré inequality in III.6 are used as stated. Their proofs, the full per-bond spectral comparison and the origin of the run’s committed-scale energy hypothesis have not been audited here.
Later applications. The dependency database identifies consumers in the units listed above. Their verification of the site-sum, per-packet and multiset hypotheses remains outside this unit’s audit. In particular, the bounds here do not construct renormalization outputs or bound the sector remainders uniformly in volume.
Unread supporting proofs. The breadth-first construction, the distinct-label forest count, the spanning-walk construction for animals and the integration of allocations restricted to sectors were used through their statements. In the auxiliary Gaussian chain, the statements and the displayed estimates of Lemma 4.36 were checked; its supporting local chart and quadratic-form comparison lemmas were not audited throughout. These remain proof inputs, rather than additional conclusions of the present sketches.
Flag locality. In III.5.9(b), locality on \(P\) applies to the raw coefficients. Their masked versions also inspect flags on \(P\); those flags can depend on plaquettes extending outside \(P\). A written proof must retain this flag dependence or enlarge the physical support before claiming ordinary link locality for the masked coefficients.
Verification scope. This is a source correspondence check. No Lean build, axiom computation or rendered-page check was run.