Towards Architectural-Level Safety Control: Relocating Hazardous Knowledge into Removable Experts in Mixture-of-Experts Models
AI & ML interests
None defined yet.
Recent Activity
Papers
ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing
A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
-
AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
Paper • 2605.29801 • Published • 144 -
AI45Research/AgentDoG1.5-Qwen3.5-4B
Text Classification • 5B • Updated • 155 • 2 -
AI45Research/AgentDoG1.5-Unified-Qwen3.5-4B
Text Classification • 4B • Updated • 78 • 3 -
AI45Research/AgentDoG1.5-FG-Qwen3.5-2B
Text Classification • 2B • Updated • 22 • 1
A Diagnostic Guardrail Framework for AI Agent Safety and Security
-
AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security
Paper • 2601.18491 • Published • 127 -
AI45Research/AgentDoG-Qwen3-4B
Text Classification • 4B • Updated • 155 • 24 -
AI45Research/ATBench
Viewer • Updated • 1.5k • 2.6k • 43 -
AI45Research/AgentDoG-FG-Llama3.1-8B
Text Classification • 8B • Updated • 34 • 9
ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
-
AI45Research/ReSI-Safin-1-35B-A3B
Text Generation • 36B • Updated • 26 • 1 -
AI45Research/ReSI-Qwen3.6-27B
Image-Text-to-Text • 27B • Updated • 22 -
AI45Research/ReSI-Ministral-3-14B-Reasoning
Image-Text-to-Text • 14B • Updated • 21 -
ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
Paper • 2610.12233 • Published • 28
SAfactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence, including open datasets and models
-
AI45Research/SafeWork-R1
Image-Text-to-Text • 73B • Updated • 31 • 6 -
AI45Research/SafeWork-R1-Qwen2.5VL-7B
Image-Text-to-Text • 8B • Updated • 21 • 1 -
AI45Research/SafeWork-R1-InternVL3-78B
Image-Text-to-Text • 78B • Updated • 10 • 1 -
AI45Research/SafeWork-R1-DeepSeek-70B
Image-Text-to-Text • 71B • Updated • 30
Towards Architectural-Level Safety Control: Relocating Hazardous Knowledge into Removable Experts in Mixture-of-Experts Models
ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
-
AI45Research/ReSI-Safin-1-35B-A3B
Text Generation • 36B • Updated • 26 • 1 -
AI45Research/ReSI-Qwen3.6-27B
Image-Text-to-Text • 27B • Updated • 22 -
AI45Research/ReSI-Ministral-3-14B-Reasoning
Image-Text-to-Text • 14B • Updated • 21 -
ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
Paper • 2610.12233 • Published • 28
A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
-
AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security
Paper • 2605.29801 • Published • 144 -
AI45Research/AgentDoG1.5-Qwen3.5-4B
Text Classification • 5B • Updated • 155 • 2 -
AI45Research/AgentDoG1.5-Unified-Qwen3.5-4B
Text Classification • 4B • Updated • 78 • 3 -
AI45Research/AgentDoG1.5-FG-Qwen3.5-2B
Text Classification • 2B • Updated • 22 • 1
SAfactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence, including open datasets and models
A Diagnostic Guardrail Framework for AI Agent Safety and Security
-
AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security
Paper • 2601.18491 • Published • 127 -
AI45Research/AgentDoG-Qwen3-4B
Text Classification • 4B • Updated • 155 • 24 -
AI45Research/ATBench
Viewer • Updated • 1.5k • 2.6k • 43 -
AI45Research/AgentDoG-FG-Llama3.1-8B
Text Classification • 8B • Updated • 34 • 9
-
AI45Research/SafeWork-R1
Image-Text-to-Text • 73B • Updated • 31 • 6 -
AI45Research/SafeWork-R1-Qwen2.5VL-7B
Image-Text-to-Text • 8B • Updated • 21 • 1 -
AI45Research/SafeWork-R1-InternVL3-78B
Image-Text-to-Text • 78B • Updated • 10 • 1 -
AI45Research/SafeWork-R1-DeepSeek-70B
Image-Text-to-Text • 71B • Updated • 30