A domain-specific multi-layer safeguard for educational LLM tutors achieves 0% false positives and 46.34% attack bypass at 2.5 ms latency on a 480-query holdout, outperforming NeMo Guardrails in usability but not fully in robustness.
Productive struggle in ai-enhanced learning environments
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs
A domain-specific multi-layer safeguard for educational LLM tutors achieves 0% false positives and 46.34% attack bypass at 2.5 ms latency on a 480-query holdout, outperforming NeMo Guardrails in usability but not fully in robustness.