Cascade-Aware Multi-Agent Routing: Adaptive Geometry Selection on Live Execution Graphs
Davide Di Gioia · 2026
Read Cascade-Aware Multi-Agent Routing: Adaptive Geometry Selection on Live Execution Graphs by Davide Di Gioia (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-17112.txt.
Description
Di Gioia studies how failures cascade through graphs of cooperating AI agents. Schedulers route tasks by load and fitness, but ignore how graph shape changes failure spread. In tree like graphs, one failure can cascade without bound. In dense cyclic graphs, it tends to self limit. He formalizes this gap as an online geometry control problem and proves a cascade sensitivity condition. His fix is a spatio temporal sidecar with two scorers, blended by a small learned gate. On 250 scenarios, the scheduler win rate rose to 87.2 percent.
Catalog record
| CREATOR | Davide Di Gioia |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57044 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.17112