Cross-modal Identity Mapping: Minimizing Information Loss in Modality Conversion via Reinforcement Learning
Paper • 2603.01696 • Published
CIM
Model weights for "Cross-modal Identity Mapping: Minimizing Information Loss in Modality Conversion via Reinforcement Learning" (CVPR 2026)
-
-
kkk5/CIM-LLaVA1.5-7B
Image-Text-to-Text • 7B • Updated • 2 -
kkk5/CIM-Qwen2-VL-7B-SFT
Image-Text-to-Text • 8B • Updated • 5 • 1 -
kkk5/CIM-Qwen2-VL-7B
Image-Text-to-Text • 8B • Updated • 3 -
kkk5/CIM-Qwen2.5-VL-7B
Image-Text-to-Text • 8B • Updated • 6 • 1 -
kkk5/CIM-InternVL2-8B
Image-Text-to-Text • 8B • Updated • 2 -
kkk5/CIM-InternVL2.5-8B
Image-Text-to-Text • 8B • Updated • 36 -
kkk5/CIM-InternVL3-8B
Image-Text-to-Text • 8B • Updated • 3 -
kkk5/CIM-LLaVA1.5-7B-SFT
Image-Text-to-Text • 7B • Updated • 4