SentGuard achieves 90.5% detection of unsafe cases within two sentences at 7.41% false positive rate by operating at sentence boundaries during LLM streaming generation.
arXiv preprint arXiv:2501.00055 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SentGuard: Sentence-Level Streaming Guardrails for Large Language Models
SentGuard achieves 90.5% detection of unsafe cases within two sentences at 7.41% false positive rate by operating at sentence boundaries during LLM streaming generation.