Saved in:
Bibliographic Details
Main Authors: Minko, Bogdan, Sadiekh, Sabrina, Kokuykin, Evgeniy
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.05277
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910195166216192
author Minko, Bogdan
Sadiekh, Sabrina
Kokuykin, Evgeniy
author_facet Minko, Bogdan
Sadiekh, Sabrina
Kokuykin, Evgeniy
contents Production LLM systems require both safety moderation and PII detection under strict latency and cost constraints. This creates a trade-off: autoregressive moderators are accurate but expensive, while lightweight encoders are faster but less capable. We present GLiNER Guard (GLiGuard), a unified encoder that performs safety classification and PII detection in a single forward pass, simplifying safety pipelines. We introduce three variants: compact uni- and bi-encoders (145-147M) for high-throughput serving, and GLiGuard Omni (209M) for stronger moderation quality. Under dynamic batching on a single A100, the compact model reaches 193 requests/sec with P99 latency below 1s, achieving 1.6x higher throughput than GLiNER2. Omni remains competitive with much larger moderators on public safety benchmarks. We also release PII-Bench, a span-level benchmark for evaluating PII detection in end-to-end pipelines. Overall, encoder-based guardrails offer a practical low-cost alternative for always-on moderation. Models and benchmarks are released on HuggingFace.
format Preprint
id arxiv_https___arxiv_org_abs_2605_05277
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle GLiNER Guard: Unified Encoder Family for Production LLM Safety and Privacy
Minko, Bogdan
Sadiekh, Sabrina
Kokuykin, Evgeniy
Cryptography and Security
Production LLM systems require both safety moderation and PII detection under strict latency and cost constraints. This creates a trade-off: autoregressive moderators are accurate but expensive, while lightweight encoders are faster but less capable. We present GLiNER Guard (GLiGuard), a unified encoder that performs safety classification and PII detection in a single forward pass, simplifying safety pipelines. We introduce three variants: compact uni- and bi-encoders (145-147M) for high-throughput serving, and GLiGuard Omni (209M) for stronger moderation quality. Under dynamic batching on a single A100, the compact model reaches 193 requests/sec with P99 latency below 1s, achieving 1.6x higher throughput than GLiNER2. Omni remains competitive with much larger moderators on public safety benchmarks. We also release PII-Bench, a span-level benchmark for evaluating PII detection in end-to-end pipelines. Overall, encoder-based guardrails offer a practical low-cost alternative for always-on moderation. Models and benchmarks are released on HuggingFace.
title GLiNER Guard: Unified Encoder Family for Production LLM Safety and Privacy
topic Cryptography and Security
url https://arxiv.org/abs/2605.05277