HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety
We present HaloGuard 1.0, an open-weights implementation of the constitutional-classifier paradigm for input safety. It achieves state-of-the-art performance on English and multilingual prompt-safety benchmarks at roughly one-tenth the model size of current leading open guard models. The safety constitution is the organising structure of the corpus: a natural-language constitution of 46 policies and 2,940 subcategories drives synthetic data generation, with exhaustive one-to-one paired counterfactuals that hold topic and vocabulary fixed while flipping intent, a two-tier harmless design that s
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 47%Nemotron 3.5 Content Safety: Customizable Multimodal Safety for Global Enterprise AI →
- PossiblePossibly related (embedding) · 47%mehaksharma2949/India_runs_data_and_ai_challenge →
- PossiblePossibly related (embedding) · 46%Transformer →
- PossiblePossibly related (embedding) · 45%AgustiPuigserver/opus-prompt-architect →
- LinkedLinked via arxiv author · 85%Navaneeth Sangameswaran →
“HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety”
- LinkedLinked via arxiv author · 85%Preetham S →
“HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety”
- LinkedLinked via arxiv author · 85%Ashmiya Lenin →
“HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety”
- PossiblePossibly related (embedding) · 52%EuroEval/EuroEval →