A 4B LLM safety guardrail trained with reasoning supervision but deployed with reasoning-free inference outperforms 8B baselines on safety benchmarks.
Shieldgemma: Generative ai content moderation based on gemma,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DT-Guard: Intent-Driven Reasoning-Active Training for Reasoning-Free LLM Safety Guardrail
A 4B LLM safety guardrail trained with reasoning supervision but deployed with reasoning-free inference outperforms 8B baselines on safety benchmarks.