Malware Detection at the Edge with Lightweight LLMs: A Performance Evaluation
Christian Rondanini 1 , Barbara Carminati 1 , Elena Ferrari 1 , Antonio Gaudiano 1, Ashish Kundu 2 · 2025
Read Malware Detection at the Edge with Lightweight LLMs: A Performance Evaluation by Christian Rondanini 1 , Barbara Carminati 1 , Elena Ferrari 1 , Antonio Gaudiano 1, Ashish Kundu 2 (2025) 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-2503-04302.txt.
Description
Christian Rondanini, Barbara Carminati, Elena Ferrari, Antonio Gaudiano, and Ashish Kundu wrote a 2025 security paper on detecting malware at the network edge with small language models. Edge devices, like sensors and gateways, have little power and memory. Traditional malware detectors struggle against modern malware's speed of change, so language model detectors are attractive if they can fit.
The authors propose an architecture that plays to small models' strengths while covering their limits. They then run a broad experimental evaluation of several state of the art lightweight models. Tests use public datasets built for edge and internet of things scenarios, with varied attack types. The study weighs accuracy against the energy and compute budgets of real edge hardware.
Catalog record
| CREATOR | Christian Rondanini 1 , Barbara Carminati 1 , Elena Ferrari 1 , Antonio Gaudiano 1, Ashish Kundu 2 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 64 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-04302