# SafeText: A Benchmark for Exploring Physical Safety in Language Models

arXiv: https://arxiv.org/abs/2210.10045 · HTML: https://mlsanity.com/paper/2210.10045/

## Key facts

- Cited by 41 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 43 references that are in the map.
- First posted on arXiv 18 Oct 2022.

## Builds on (most cited first, 25 of 25)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [A large annotated corpus for learning natural language inference](https://mlsanity.com/paper/1508.05326/) (2015)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [FEVER: a large-scale dataset for Fact Extraction and VERification](https://mlsanity.com/paper/1803.05355/) (2018)
- [Ethical and social risks of harm from Language Models](https://mlsanity.com/paper/2112.04359/) (2021)
- [Aligning AI With Shared Human Values](https://mlsanity.com/paper/2008.02275/) (2020)
- [From Recognition to Cognition: Visual Commonsense Reasoning](https://mlsanity.com/paper/1811.10830/) (2018)
- [COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge Graphs](https://mlsanity.com/paper/2010.05953/) (2020)
- [Latent Hatred: A Benchmark for Understanding Implicit Hate Speech](https://mlsanity.com/paper/2109.05322/) (2021)
- [On Hallucination and Predictive Uncertainty in Conditional Language Generation](https://mlsanity.com/paper/2103.15025/) (2021)
- [Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text](https://mlsanity.com/paper/2202.06935/) (2022)
- [The Moral Integrity Corpus: A Benchmark for Ethical Dialogue Systems](https://mlsanity.com/paper/2204.03021/) (2022)
- [On the Safety of Conversational Models: Taxonomy, Dataset, and Benchmark](https://mlsanity.com/paper/2110.08466/) (2021)
- [Ethical-Advice Taker: Do Language Models Understand Natural Language Interventions?](https://mlsanity.com/paper/2106.01465/) (2021)
- [ANLIzing the Adversarial Natural Language Inference Dataset](https://mlsanity.com/paper/2010.12729/) (2020)
- [Mitigating Covertly Unsafe Text within Natural Language Systems](https://mlsanity.com/paper/2210.09306/) (2022)
- [Detecting Gang-Involved Escalation on Social Media Using Context](https://mlsanity.com/paper/1809.03632/) (2018)
- [SkillBot: Identifying Risky Content for Children in Alexa Skills](https://mlsanity.com/paper/2102.03382/) (2021)
- ['Beach' to 'Bitch': Inadvertent Unsafe Transcription of Kids' Content on YouTube](https://mlsanity.com/paper/2203.04837/) (2022)

## Built on by (most cited first, 30 of 41)

- [XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models](https://mlsanity.com/paper/2308.01263/) (2023)
- [Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions](https://mlsanity.com/paper/2309.07875/) (2023)
- [Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models](https://mlsanity.com/paper/2310.02949/) (2023)
- [OR-Bench: An Over-Refusal Benchmark for Large Language Models](https://mlsanity.com/paper/2405.20947/) (2024)
- [Automatically Correcting Large Language Models: Surveying the landscape of diverse self-correction strategies](https://mlsanity.com/paper/2308.03188/) (2023)
- [SafetyBench: Evaluating the Safety of Large Language Models](https://mlsanity.com/paper/2309.07045/) (2023)
- [SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents](https://mlsanity.com/paper/2412.13178/) (2024)
- [Language Model Behavior: A Comprehensive Survey](https://mlsanity.com/paper/2303.11504/) (2023)
- [SafetyPrompts: a Systematic Review of Open Datasets for Evaluating and Improving Large Language Model Safety](https://mlsanity.com/paper/2404.05399/) (2024)
- [All Languages Matter: On the Multilingual Safety of Large Language Models](https://mlsanity.com/paper/2310.00905/) (2023)
- [Value FULCRA: Mapping Large Language Models to the Multidimensional Spectrum of Basic Human Values](https://mlsanity.com/paper/2311.10766/) (2023)
- [Training Data Extraction From Pre-trained Language Models: A Survey](https://mlsanity.com/paper/2305.16157/) (2023)
- [From Instructions to Intrinsic Human Values -- A Survey of Alignment Goals for Big Models](https://mlsanity.com/paper/2308.12014/) (2023)
- [RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models](https://mlsanity.com/paper/2504.18041/) (2025)
- [Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs](https://mlsanity.com/paper/2502.11184/) (2025)
- [Investigating the Impact of Quantization Methods on the Safety and Reliability of Large Language Models](https://mlsanity.com/paper/2502.15799/) (2025)
- [Towards Safer Generative Language Models: A Survey on Safety Risks, Evaluations, and Improvements](https://mlsanity.com/paper/2302.09270/) (2023)
- [Building Trustworthy NeuroSymbolic AI Systems: Consistency, Reliability, Explainability, and Safety](https://mlsanity.com/paper/2312.06798/) (2023)
- [ASSERT: Automated Safety Scenario Red Teaming for Evaluating the Robustness of Large Language Models](https://mlsanity.com/paper/2310.09624/) (2023)
- [CRoW: Benchmarking Commonsense Reasoning in Real-World Tasks](https://mlsanity.com/paper/2310.15239/) (2023)
- [FLEX: A Benchmark for Evaluating Robustness of Fairness in Large Language Models](https://mlsanity.com/paper/2503.19540/) (2025)
- [IndoSafety: Culturally Grounded Safety for LLMs in Indonesian Languages](https://mlsanity.com/paper/2506.02573/) (2025)
- [Similarity-weighted Construction of Contextualized Commonsense Knowledge Graphs for Knowledge-intense Argumentation Tasks](https://mlsanity.com/paper/2305.08495/) (2023)
- [Towards Evaluating Proactive Risk Awareness of Multimodal Language Models](https://mlsanity.com/paper/2505.17455/) (2025)
- [Foveate, Attribute, and Rationalize: Towards Physically Safe and Trustworthy AI](https://mlsanity.com/paper/2212.09667/) (2022)
- [Enhancing Safety in Reinforcement Learning with Human Feedback via Rectified Policy Optimization](https://mlsanity.com/paper/2410.19933/) (2024)
- [Evaluating and Improving Robustness in Large Language Models: A Survey and Future Directions](https://mlsanity.com/paper/2506.11111/) (2025)
- [SafeLLM: Domain-Specific Safety Monitoring for Large Language Models: A Case Study of Offshore Wind Maintenance](https://mlsanity.com/paper/2410.10852/) (2024)
- [Attribute Diversity Determines the Systematicity Gap in VQA](https://mlsanity.com/paper/2311.08695/) (2023)
- [CHBench: A Chinese Dataset for Evaluating Health in Large Language Models](https://mlsanity.com/paper/2409.15766/) (2024)

## Abstract (arXiv)

Understanding what constitutes safe text is an important issue in natural language processing and can often prevent the deployment of models deemed harmful and unsafe. One such type of safety that has been scarcely studied is commonsense physical safety, i.e. text that is not explicitly violent and requires additional commonsense knowledge to comprehend that it leads to physical harm. We create the first benchmark dataset, SafeText, comprising real-life scenarios with paired safe and physically unsafe pieces of advice. We utilize SafeText to empirically study commonsense physical safety across various models designed for text generation and commonsense reasoning tasks. We find that state-of-the-art large language models are susceptible to the generation of unsafe text and have difficulty rejecting unsafe advice. As a result, we argue for further studies of safety and the assessment of commonsense physical safety in models before release.