Brief

A new system called GLiNER Guard (GLiGuard) has been developed to streamline safety moderation and PII detection for large language models. This unified encoder collapses multiple classifiers and NER models into a single forward pass, significantly reducing processing time and cost compared to existing autoregressive or fragmented encoder approaches. GLiGuard's schema-driven interface allows for dynamic policy changes without retraining, making it a more efficient solution for production LLM applications. AI

Researchers have developed a new method called CPD Online to detect adversarial prompts that attempt to jailbreak large language models. This technique treats prompt detection as an online change-point detection problem, analyzing sequential entropy changes in the model's token predictions. CPD Online is model-agnostic, requires no training, and can pinpoint the onset of malicious prompts, outperforming existing perplexity-based detectors on various open-weight models. AI