RoMA, a statistical robustness estimator, is adapted to black-box language models and is shown to approximate exact verification within 1% on small networks while scaling to BERT sentiment analysis.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Statistical Runtime Verification for LLMs via Robustness Estimation
RoMA, a statistical robustness estimator, is adapted to black-box language models and is shown to approximate exact verification within 1% on small networks while scaling to BERT sentiment analysis.