LionGuard: Building a Contextualized Moderation Classifier to Tackle Localized Unsafe Content
Paper
• 2407.10995 • Published
• 2
Papers by GovTech 📝
-
-
A Flexible Large Language Models Guardrail Development Methodology Applied to Off-Topic Prompt Detection
Paper • 2411.12946 • Published • 22 -
Safe at the Margins: A General Approach to Safety Alignment in Low-Resource English Languages -- A Singlish Case Study
Paper • 2502.12485 • Published • 2 -
MinorBench: A hand-built benchmark for content-based risks for children
Paper • 2503.10242 • Published • 5 -
Know Or Not: a library for evaluating out-of-knowledge base robustness
Paper • 2505.13545 • Published -
RabakBench: Scaling Human Annotations to Construct Localized Multilingual Safety Benchmarks for Low-Resource Languages
Paper • 2507.05980 • Published • 2 -
Measuring What Matters: A Framework for Evaluating Safety Risks in Real-World LLM Applications
Paper • 2507.09820 • Published -
Toxicity-Aware Few-Shot Prompting for Low-Resource Singlish Translation
Paper • 2507.11966 • Published -
LionGuard 2: Building Lightweight, Data-Efficient & Localised Multilingual Content Moderators
Paper • 2507.15339 • Published • 1 -
Running in CIRCLE? A Simple Benchmark for LLM Code Interpreter Security
Paper • 2507.19399 • Published • 2 -
Reasoning Beyond the Obvious: Evaluating Divergent and Convergent Thinking in LLMs for Financial Scenarios
Paper • 2507.18368 • Published • 1