Qwen3-4B-SafeRL is a safety-aligned version of the Qwen3-4B model. It enhances the model's robustness against harmful or adversarial prompts through reinforcement learning training. This version uses a hybrid reward function for optimization, balancing three objectives: safety, usefulness, and minimizing unnecessary rejections.

Natural Language ProcessingTransformersEnglish

Qwen3-4B-SafeRL is a safety-aligned version based on the Qwen3-4B model. It is trained through reinforcement learning and combined with the reward signals of Qwen3Guard-Gen, enhancing the model's robustness against harmful or adversarial prompts. While ensuring safety, it avoids overly simple or evasive rejection behaviors.

Natural Language ProcessingTransformers

A target detection model based on YOLO11, specifically designed for construction site safety monitoring. It can detect various safety hazards such as workers not wearing safety equipment, intrusion into dangerous areas, and equipment approaching hazard sources.

Computer VisionOnnxEnglish

M-Hood is a series of models specifically designed to mitigate the hallucination phenomenon in object detection. Through novel fine-tuning strategies and a revised benchmark dataset, it significantly reduces false alarms on out-of-distribution data and enhances the safety and reliability of object detection systems.

Computer VisionPytorch

Owlet Safety 1 is a multi-label safety event detection model fine-tuned based on Qwen2.5-VL-3B-Instruct. It is specifically designed for safety activity recognition in video surveillance and can simultaneously detect multiple safety-related events such as fires, smoke, falls, and assaults.

MultimodalTransformers

DMind-1 is a Web3 expert model built upon Qwen3-32B, optimized for the Web3 ecosystem through supervised instruction fine-tuning and human feedback reinforcement learning, achieving significant improvements in task accuracy, content safety, and expert-level interaction alignment.

Natural Language ProcessingTransformersMultiple Languages

QwenGuard is a visual security protection model capable of evaluating images based on provided security policies, outputting safety ratings, safety categories, and evaluation rationale.

MultimodalTransformers

QwenGuard-v1.2-3B is a visual security protection model developed based on Qwen/Qwen2.5-VL-3B-Instruct, designed for assessing the safety of image content.

MultimodalTransformers

This model is specifically fine-tuned for NSFW image classification, categorizing content into three safety-critical classes, suitable for content moderation, safety filtering, and compliant content processing systems.

Computer VisionTransformersEnglish

Llama Guard 4 is a native multimodal safety classifier with 12 billion parameters, jointly trained on text and multiple images for content safety evaluation of large language model inputs and outputs.

MultimodalTransformersEnglish

Beaver-7B-v3.0 is a 7B-parameter large language model based on the LLaMA architecture, focusing on safety and human feedback reinforcement learning (RLHF).

Natural Language ProcessingGgufEnglish

A medical text safety and ethics classifier based on the ModernBERT architecture, specifically designed for evaluating text safety in the medical field

Natural Language ProcessingSafetensorsEnglish

ShieldGemma 2 is a model trained on Gemma 3's 4-billion-parameter IT checkpoint for cross-critical-category image safety classification, receiving images and outputting policy-compliant safety labels.

MultimodalTransformers

Phi-4-mini-instruct is a lightweight open-source model built on synthetic data and curated public websites, focusing on high-quality data with strong reasoning capabilities. It supports a 128K token context length and is enhanced through supervised fine-tuning and direct preference optimization to ensure precise instruction following and safety.

Natural Language ProcessingTransformersMultiple Languages

ThinkGuard is an advanced guardrail model designed to enhance safety classification capabilities through deliberate slow thinking.

Natural Language ProcessingTransformersEnglish

PolyGuard is a multilingual safety model for protecting LLM-generated content, supporting safety review in 17 languages, currently achieving state-of-the-art performance.

Natural Language ProcessingSafetensorsMultiple Languages

DuoGuard-0.5B is a multilingual, decoder-only large language model-based classifier, specifically designed for safety content moderation across 12 different subcategories.

Natural Language ProcessingTransformers

GuardReasoner 8B is a fine-tuned model based on meta-llama/Llama-3.1-8B, specializing in reasoning-based LLM safety protection

Natural Language ProcessingTransformers

An 8-billion-parameter brand safety classification model optimized based on Llama 3.1, specifically designed for Apple chip devices

Natural Language ProcessingTransformersMultiple Languages

VISION-1 is a content moderation and safety classification model based on transformers, focusing on text classification tasks.

Natural Language ProcessingGgufEnglish