The Setting
Take a critical Galton-Watson tree with an even offspring distribution conditioned, for a size parameter , to have vertices, together with vertex marks drawn independently given the vertex degrees. Summing a bounded function of each ancestral step and its mark along the path from the root to a vertex yields a spine sum whose slope per unit height is governed by a stationary spine expectation. The question is how the maximal deviation across all vertices between this path sum and the corresponding multiple of vertex height behaves at scale as .
Tree Codes and Conditioned Offspring Laws
Let , and let be a finite degree sequence. For any step , the degree sequence partial walk is given by (1): (1)✦ A discrete bridge sequence of length is a sequence in satisfying .✦ A sequence satisfies the valid tree code condition if and for all .✦ A Lukasiewicz tree code of length is a sequence in satisfying and for all , and is the set of Lukasiewicz tree codes of length .✦ For any and , the tree vertex height is the cardinality in (2): (2)✦
For and an offset , the cyclic degree rotation sends to ✦, and the cyclic bridge rotation restricts to discrete bridge sequences of length .✦ For each discrete bridge sequence of positive length , the good rotation index is the unique cyclic shift index whose rotated bridge satisfies the valid tree code condition✦, and the canonical tree code of is its cyclic degree rotation by .✦
A critical even offspring distribution specifies a nonnegative mass function on satisfying , for every odd , , , , and .✦ For a critical even offspring distribution and , the independent and identically distributed (i.i.d.) offspring product measure of on is the -fold product measure of .✦ For and a sequence , the total step of is .✦ For a size parameter , the conditioned offspring bridge law on is the conditional distribution of given the event .✦ Let be the tree-code map from bridges on sequences of length ; the conditioned tree code law on is the pushforward of under .✦
The Offspring Spine Ladder Distribution
The infinite offspring product measure of on is the countable product of copies of the offspring law of indexed by .✦ For a sequence and a step , the offspring walk partial sum of at step is given by (3): (3)✦ An index satisfies the ladder epoch condition for if either , or for all and .✦ The first weak ladder epoch of is the least index satisfying the ladder epoch condition for ✦, and the weak ladder overshoot of is .✦ Together these form the spine ladder variable pair ✦, and the offspring spine ladder distribution on is the pushforward of under .✦
Spine Records and Marked Sums
For and a sequence , the zero-extended sequence is defined by for and for .✦ For , the reversed prefix index is if and if ✦; for and , the reversed prefix of of length has entries .✦
For an integer sequence and , the running maximum of up to is .✦ For a sequence taking values in a linearly ordered set and , the set of weak record times of up to is given by (4): (4)✦ Writing , for a sequence and a step , the record type of at step is defined in (5) by (5)✦ For a weight function , a sequence , and a step bound , the record spine sum is defined in (6) by (6)✦ For , a Lukasiewicz tree code , a weight function , and a step bound , the tree code spine sum of is .✦
For and , the spine records of at are the elements ✦, and the spine record type of is .✦ When , each spine record determines the spine record vertex .✦ Given a mark set , a function , , , and a mark assignment , the marked spine sum is given by (7): (7)✦
Let be a measurable space and let be a family of probability measures on indexed by . For and a Lukasiewicz tree code , the conditional mark measure on is the product measure ✦, and the map defines the tree code marking kernel from to .✦ At size parameter , the marked tree code law on is the joint distribution of the conditioned tree code law with the tree code marking kernel .✦
The Claim
Theorem 1. Let be a measurable space with a probability measure under which is a standard Brownian motion, let be a critical even offspring distribution, and let and . For every function satisfying for all , writing , the tree code spine sum satisfies (8): (8)✦ Moreover, for every measurable space , every family of probability measures on indexed by , and every function such that is measurable for each and for all and , writing, for each , , and writing , the marked spine sum satisfies (9) and the mark noise satisfies (10): (9)✦ (10)✦
In the paper, the result reads: for every fixed bounded measurable real label , , where , and , , and stand for , , and . The theorem expresses this convergence in probability on conditioned Lukasiewicz tree codes of length through vanishing tail probabilities at threshold for the unweighted spine sum in (8), the marked spine sum in (9), and the mark noise in (10).
What Is Checked
The claims are stated over a probability space supporting a standard Brownian motion , a critical even offspring distribution , its conditioned tree code law on , its marked tree code law on built from a family of probability measures on a measurable space , the offspring spine ladder distribution , the tree vertex height, the tree code spine sum, the marked spine sum, bounded weights and measurable mark weights bounded in absolute value by , and a threshold . For these constructions, the library proves that cyclic degree and bridge rotations act as permutations on and on discrete bridge sequences of length , that a discrete bridge sequence of length rotated by its good rotation index satisfies the valid tree code condition✦, and that every sequence in admits an index satisfying the ladder epoch condition.✦