Shieldstral
We introduce Shieldstral, a 3B-parameter policy-adaptive multimodal safety classifier that matches or outperforms models nearly 7times its size on text safety benchmarks and sets a…
Hugging Face · Daily Papers
·Antonia Calvi, Avinash Sooriyarachchi
·
·▲ 7 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Antonia Calvi, Avinash Sooriyarachchi, Giada Pistilli, Guillaume Lample, Maarten Buyl, Maximilian Augustin
- 7 upvotes da comunidade
Resumo
Resumo original (em inglês), extraído do paper:
We introduce Shieldstral, a 3B-parameter policy-adaptive multimodal safety classifier that matches or outperforms models nearly 7times its size on text safety benchmarks and sets a new state of the art on multimodal safety classification. Shieldstral formulates content moderation as a binary question-answering task. This simple formulation unifies diverse moderation tasks into a single yes/no problem, enabling heterogeneous safety datasets with divergent taxonomies to be consolidated under one training framework. We present the data construction recipe, covering curation and generation of approximately 54.1M samples and a fine-grained evaluation set to evaluate policy adaptability. Together, these enable a small adaptive model to match or outperform much larger models.Onde ler
// relacionados
Leia também
Blog
Liquid AI lança LFM2.5-Encoder-230M e LFM2.5-Encoder-350M: codificadores bidirecionais que permanecem rápidos com contexto de 8K na CPU
Blog
Memory for Large Language Models
Blog
When Does Deep Representation Learning Help Single-Cell Clustering? A Sensitivity-Aware Diagnostic Benchmark for Biomedical AI Pipelines
Blog