ChatGPT-4o, prompted with prior knowledge about normal battery temperature patterns, detects battery thermal anomalies at 86.6% AUC without training, though most trained methods score higher.
Exploiting autoen- coder’s weakness to generate pseudo anomalies,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Zero-Shot Anomaly Detection in Battery Thermal Images Using Visual Question Answering with Prior Knowledge
ChatGPT-4o, prompted with prior knowledge about normal battery temperature patterns, detects battery thermal anomalies at 86.6% AUC without training, though most trained methods score higher.