Mistral releases Shieldstral, a 3B open multimodal safety classifier
Mistral AI released Shieldstral, a 3B-parameter multimodal safety classifier under Apache 2.0 that it says matches text-safety performance of models up to about 7x larger. The lightweight open-weight filter is aimed at adding safety checks without the cost of much larger guardrail models.
LLMgram editorial selection and synthesis · @llmgram. LLMgram is not the original publisher of this information.
Continue on LLMgram: Open in AI Signal →
Original reporting: Mistral releases Shieldstral, a 3B multimodal safety classifier that it says matches models up to 7x its size on text safety, available under Apache 2.0 (Mistral AI Blog)