Fuzzing-Gestützte Autonome Systeme
Anmelder: BAE SYSTEMS plc 🇬🇧
Details
- Veröffentlichungs-Nr.
- EP4745784
- Anmeldetag
- 14. November 2024
- Veröffentlichung
- 20. Mai 2026
- Rechtsraum
- EP
- IPC
- G06F11/3668, // G06N20/00
Abstract
A computer-implemented method is disclosed, in which a Large Language Model, LLM, is used to score or rank a plurality of test cases for use in fuzz testing an autonomous system. The method comprises: receiving a definition of interestingness and a plurality of test cases; providing the definition of interestingness and the plurality of test cases to the LLM in the form of first input text and second input text, respectively; and receiving an output from the LLM. The output is indicative of a respective score or ranking assigned to each one of the plurality of test cases by the LLM in accordance with the definition of interestingness. The autonomous system may then be fuzz tested by selecting one or more of the plurality of test cases in accordance with the assigned scores or rankings, and executing the selected test case(s) by the autonomous system.
Anmelder
- Firma
- BAE SYSTEMS plc
- Land
- 🇬🇧 Großbritannien
Britisches Rüstungs- und Luftfahrtunternehmen mit Sitz in London. BAE Systems entwickelt Verteidigungstechnik, Luftfahrzeuge, Marineschiffe sowie Sicherheits- und Elektroniksysteme.
1.114 Patente in unserer Datenbank
Vertreten von
-
BAE SYSTEMS plc Group IP Department