Library
85,762 resources from 11 public sources. 54 checked by hand; the rest are unreviewed.
Results
-
PKU-SafeRLHF
Datasets & benchmarksHugging Face dataset by PKU-Alignment
PKU-Alignment/PKU-SafeRLHF -
relaion2B-en-research-safe
Datasets & benchmarksHugging Face dataset by laion
laion/relaion2B-en-research-safe -
Omni-SafetyBench
Datasets & benchmarks[ACM MM 2026 Oral] Code for paper "Omni-SafetyBench: A Benchmark for Safety Evaluation of Audio-Visual Large Language Models".
git clone https://github.com/THU-BPM/Omni-SafetyBench.git -
wildjailbreak
Datasets & benchmarksHugging Face dataset by allenai
allenai/wildjailbreak -
autoguardrails
Datasets & benchmarksAlignment-research scaffold (autoresearch-style) for LLM guardrails: search over a single policy.md surface
git clone https://github.com/SantanderAI/autoguardrails.git -
ai-village
Datasets & benchmarksHugging Face dataset by aidigestorg
aidigestorg/ai-village -
TrustLLM
Datasets & benchmarks[ICML 2024] TrustLLM: LLM trustworthiness evaluation across truthfulness, safety, fairness, robustness, privacy and ethics. Python/CLI toolkit for local models, compatible APIs and AI agent orchestration.
git clone https://github.com/HowieHwong/TrustLLM.git -
chembench
Datasets & benchmarksHow good are LLMs at chemistry?
git clone https://github.com/lamalab-org/chembench.git -
Pliny_HackAPrompt_Dataset
Datasets & benchmarksHugging Face dataset by hackaprompt
hackaprompt/Pliny_HackAPrompt_Dataset -
opencompass
Datasets & benchmarksOpenCompass is an LLM evaluation platform, supporting a wide range of models from OpenAI, Anthropic, Gemini, Qwen, GLM, DeepSeek, etc, across 100+ datasets covering knowledge, reasoning, coding, science, language, long-context, and safety.
git clone https://github.com/open-compass/opencompass.git -
JBB-Behaviors
Datasets & benchmarksHugging Face dataset by JailbreakBench
JailbreakBench/JBB-Behaviors -
sacp
Datasets & benchmarksState-Aware Collaboration Protocol: auditable work receipts for AI agents.
git clone https://github.com/aDragon0707/sacp.git
Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.