XuankunRong/SafeTag-VL-3K
SafeTag-VL-3K Dataset The SafeTag-VL-3K dataset is a multimodal dataset constructed for self-rewarded multimodal safety alignment, as presented in the paper SafeGRPO: Self-Rewarded Multimodal Safety Alignment via Rule-Governed Policy Optimization. This dataset features explicit visual, textual, and combined safety tags, which are crucial for enhancing multimodal large language models' (MLLMs) safety awareness, compositional robustness, and reasoning stability against complex… See the full description on the dataset page: https://huggingface.co/datasets/XuankunRong/SafeTag-VL-3K.
| Author / organization | XuankunRong |
|---|---|
| Downloads | 145 (as reported by Hugging Face) |
| Likes | 3 |
| Created | 31 October 2025 |
| Last updated at source | 24 February 2026 |
Popularity, last 2 days
145 now (-1 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
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 Information may be incomplete or out of date.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.