Refutations of pebble minimization via output languages

TL;DR

This paper investigates the limitations of pebble transducers in computing certain polyregular functions, demonstrating that some quadratic functions require more pebbles than previously thought and refuting related conjectures.

Contribution

The authors provide two simpler proofs showing that some quadratic polyregular functions need more pebbles than their output length suggests, and disprove a conjectured logical characterization.

Findings

Some quadratic polyregular functions require 3 pebbles.

Existence of polyregular functions with quadratic growth not computable by k-pebble transducers.

Refutation of a conjectured logical characterization of polyblind functions.

Abstract

Polyregular functions are the class of string-to-string functions definable by pebble transducers, an extension of finite-state automata with outputs and multiple two-way reading heads (pebbles) with a stack discipline. If a polyregular function can be computed with $k$ pebbles, then its output length is bounded by a polynomial of degree $k$ in the input length. But Boja\'nczyk has shown that the converse fails. In this paper, we provide two alternative easier proofs. The first establishes by elementary means that some quadratic polyregular function requires 3 pebbles. The second proof - just as short, albeit less elementary - shows a stronger statement: for every $k$, there exists some polyregular function with quadratic growth whose output language differs from that of any $k$-fold composition of macro tree transducers (and which therefore cannot be computed by a $k$-pebble transducer). Along the way, we also refute a conjectured logical characterization of polyblind functions.