A behavioral monitoring technique using HTTP, lexical, and timing signals detects guardrail presence with 100% accuracy and distinguishes guardrail blocks from LLM rejections with 98% average F1 on unseen prompts.
AEGIS 2.0: A Diverse AI Safety Dataset and Risks Taxonomy for Alignment of LLM Guardrails
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
verdicts
UNVERDICTED 8roles
dataset 1polarities
use dataset 1representative citing papers
Guardrail classifiers receive formal guarantees by certifying convex harmful regions in pre-activation space, exposing safety holes in three toxicity models despite high empirical scores.
BELLS-O is the first vendor-neutral operational benchmark comparing specialized guardrails and repurposed frontier LLMs on accuracy, false-positive rates, speed, and monetary cost across 11 harm categories and 13 jailbreak techniques.
kNNGuard classifies prompts using multi-layer kNN on LLM hidden activations from 50 examples, matching or exceeding fine-tuned guardrails in F1 while running 2.7x to 10x faster with no training required.
CDR-Bench shows state-of-the-art LLMs fail at compositional and especially order-sensitive data refinement across atomic, order-agnostic, and order-sensitive settings.
Guardian-as-an-Advisor prepends risk labels and explanations from a guardian model to queries, improving LLM safety compliance and reducing over-refusal while adding minimal compute overhead.
NVIDIA releases the Nemotron 3 model family with hybrid Mamba-Transformer architecture, LatentMoE, NVFP4 training, MTP layers, and multi-environment RL post-training for reasoning and agentic tasks.
TWGuard achieves +0.289 F1 improvement and 94.9% false-positive reduction for LLM safety guardrails in the Taiwan linguistic context compared to foundation models and baselines.
citing papers explorer
-
CDR-Bench: Evaluating Faithful Execution of Compositional, Order-Sensitive Data Refinement Recipes
CDR-Bench shows state-of-the-art LLMs fail at compositional and especially order-sensitive data refinement across atomic, order-agnostic, and order-sensitive settings.