walledai/SGXSTest
Abstract For testing refusal behavior in a cultural setting, we introduce SGXSTEST — a set of manually curated prompts designed to measure exaggerated safety within the context of Singaporean culture. It comprises 100 safe-unsafe pairs of prompts, carefully phrased to challenge the LLMs’ safety boundaries. The dataset covers 10 categories of hazards (adapted from Röttger et al. (2023)), with 10 safe-unsafe prompt pairs in each category. These categories include homonyms… See the full description on the dataset page: https://huggingface.co/datasets/walledai/SGXSTest.
| Author / organization | walledai |
|---|---|
| License (as declared) | apache-2.0 |
| Downloads | 157 (as reported by Hugging Face) |
| Likes | 3 |
| Created | 4 August 2024 |
| Last updated at source | 11 August 2024 |
Popularity, last 2 days
157 now (-6 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 17 Oct 2026 Description: Abstract For testing refusal behavior in a cultural setting, we introduce SGXSTEST — a set of manually curated prompts designed to measure exaggerated safety within the context of Singaporean culture. It comprises 100 safe-unsafe pairs of p → Abstract For testing refusal behavior in a cultural setting, we introduce SGXSTEST — a set of manually curated prompts designed to measure exaggerated safety within the context of Singaporean culture. It comprises 100 safe-unsafe pairs of p
Changes detected by Silytics between refreshes of the source.
More from walledai
- walledai/AdvBenchDataset
- walledai/XSTestDataset
- walledai/HarmBenchDataset
- walledai/StrongREJECTDataset
- walledai/BBQDataset
- walledai/MaliciousInstructDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.