CLAIAug 7, 2024

WalledEval: A Comprehensive Safety Evaluation Toolkit for Large Language Models

arXiv:2408.03837v332 citationsh-index: 77Has Code
Originality Synthesis-oriented
AI Analysis

This provides a comprehensive safety evaluation framework for AI developers and researchers, though it is incremental as it builds on existing safety testing approaches.

The researchers tackled the problem of evaluating safety in large language models by developing WalledEval, a toolkit with over 35 benchmarks and custom mutators, which includes a new content moderation tool and datasets for assessing exaggerated safety.

WalledEval is a comprehensive AI safety testing toolkit designed to evaluate large language models (LLMs). It accommodates a diverse range of models, including both open-weight and API-based ones, and features over 35 safety benchmarks covering areas such as multilingual safety, exaggerated safety, and prompt injections. The framework supports both LLM and judge benchmarking and incorporates custom mutators to test safety against various text-style mutations, such as future tense and paraphrasing. Additionally, WalledEval introduces WalledGuard, a new, small, and performant content moderation tool, and two datasets: SGXSTest and HIXSTest, which serve as benchmarks for assessing the exaggerated safety of LLMs and judges in cultural contexts. We make WalledEval publicly available at https://github.com/walledai/walledeval.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes