Rethinking Positional Encoding for Neural Vehicle Routing

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hua, Chuanbo, Berto, Federico, Hottung, Andre, Zepeda, Nayeli Gast, Ma, Yining, Ma, Zihan, Wong-Chung, Paula, Kwon, Changhyun, Wu, Cathy, Tierney, Kevin, Park, Jinkyoo
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911673858654208
author Hua, Chuanbo
Berto, Federico
Hottung, Andre
Zepeda, Nayeli Gast
Ma, Yining
Ma, Zihan
Wong-Chung, Paula
Kwon, Changhyun
Wu, Cathy
Tierney, Kevin
Park, Jinkyoo
author_facet Hua, Chuanbo
Berto, Federico
Hottung, Andre
Zepeda, Nayeli Gast
Ma, Yining
Ma, Zihan
Wong-Chung, Paula
Kwon, Changhyun
Wu, Cathy
Tierney, Kevin
Park, Jinkyoo
contents Transformer-based models have become the dominant paradigm for neural combinatorial optimization (NCO) of vehicle routing problems (VRPs), yet the role of positional encoding (PE) in these architectures remains largely unexplored. Unlike natural language, where tokens are uniformly spaced on a line, routing solutions exhibit several properties that render standard NLP positional encodings inadequate. In this work, we formalize three such structural properties that a routing-aware PE should respect, namely anisometric node distances, cyclic and direction-aware topology, and hierarchical depot-anchored global multi-route structure, combining them with a unifying design principle of geometric grounding. Guided by these criteria, we analyze and compare PE methods spanning NLP, graph-transformer, and routing-specific families, and propose a hierarchical anisometric PE that combines a distance-indexed, circularly consistent in-route encoding with a depot-anchored angular cross-route encoding. Extensive experiments across diverse VRP variants demonstrate that geometry-grounded PE consistently outperforms index-based alternatives, with gains that transfer across problem variants, model architectures, and distribution shifts.
format Preprint
id arxiv_https___arxiv_org_abs_2605_11910
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Rethinking Positional Encoding for Neural Vehicle Routing
Hua, Chuanbo
Berto, Federico
Hottung, Andre
Zepeda, Nayeli Gast
Ma, Yining
Ma, Zihan
Wong-Chung, Paula
Kwon, Changhyun
Wu, Cathy
Tierney, Kevin
Park, Jinkyoo
Artificial Intelligence
Transformer-based models have become the dominant paradigm for neural combinatorial optimization (NCO) of vehicle routing problems (VRPs), yet the role of positional encoding (PE) in these architectures remains largely unexplored. Unlike natural language, where tokens are uniformly spaced on a line, routing solutions exhibit several properties that render standard NLP positional encodings inadequate. In this work, we formalize three such structural properties that a routing-aware PE should respect, namely anisometric node distances, cyclic and direction-aware topology, and hierarchical depot-anchored global multi-route structure, combining them with a unifying design principle of geometric grounding. Guided by these criteria, we analyze and compare PE methods spanning NLP, graph-transformer, and routing-specific families, and propose a hierarchical anisometric PE that combines a distance-indexed, circularly consistent in-route encoding with a depot-anchored angular cross-route encoding. Extensive experiments across diverse VRP variants demonstrate that geometry-grounded PE consistently outperforms index-based alternatives, with gains that transfer across problem variants, model architectures, and distribution shifts.
title Rethinking Positional Encoding for Neural Vehicle Routing
topic Artificial Intelligence
url https://arxiv.org/abs/2605.11910