Where-OPD: Spatially Guided On-Policy Self-Distillation of MLLMs with Synthetic Scenes
cs.CV, cs.AI, cs.CL, cs.LG
Submitted: 2026-10-01
Updated: 2026-10-01
Terminology
Sources
- OPD-V: Visual On-Policy Self-Distillation with Modality Balance
- Rethinking the Mixture of Vision Encoders Paradigm for Enhanced Visual Understanding in Multimodal LLMs
- Qwen3-VL Technical Report
- Seeing Beyond Words: Self-Supervised Visual Learning for Multimodal Large Language Models
- Thinking Without Images: Internalizing Visual Manipulation with On-Policy Self-Distillation
- Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic
- Hidden in plain sight: VLMs overlook their visual representations
- Visual Contrastive Self-Distillation
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders
- Boosting Visual Instruction Tuning with Self-Supervised Guidance
- CountQA: How Well Do MLLMs Count in the Wild?
- ViCuR: Visual Cues as Recoverable Privilege for Multimodal On-Policy Distillation
- AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation
- Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
- Visual Representation Alignment for Multimodal Large Language Models
- Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation
- Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
- RP-OPSD: Resolution-Privileged On-Policy Self-Distillation for Multimodal Large Language Models
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models