lmarena-ai/Arena-T2I-Training
Arena-T2I-Training-1K This release contains 992 training prompts for the paper Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards. Files train.json: prompt records and rubric metadata. train_hack_labels.json: prompt-specific reward-hack applicability labels. train_render_text.json: prompt-specific strings relevant to text-rendering checks. train.json record keys prompt_id: ID (0–991) prompt: text-to-image… See the full description on the dataset page: https://huggingface.co/datasets/lmarena-ai/Arena-T2I-Training.
| Author / organization | lmarena-ai |
|---|---|
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 21 October 2026 |
| Last updated at source | 22 October 2026 |
Recent changes
- 22 Oct 2026 License: mit → none
- 22 Oct 2026 Description: none → Arena-T2I-Training-1K This release contains 992 training prompts for the paper Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards. Files train.json: prompt records and rubric metadata. train_hack_labels.j
Changes detected by Silytics between refreshes of the source.
More from lmarena-ai
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 22 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.