Evidence-Aligned Multimodal On-Policy Self-Distillation for Fine-Grained Visual Understanding

arXiv:2609.34672 · cs.CV, cs.AI · Submitted 2026-09-28 · Read on arXiv

cs.CV, cs.AI

Submitted: 2026-09-28

Updated: 2026-09-28

Terminology

• Distillation — 43× in this paper · explained in 2 episodes
• Benchmark — 23× in this paper · explained in 2 episodes
• Self-distillation — 15× in this paper · explained in 6 episodes
• On-Policy Self-Distillation — 13× in this paper
• Accuracy — 11× in this paper · explained in 1 episode
• Alignment — 10× in this paper · explained in 1 episode
• Cosine Similarity — 8× in this paper · explained in 3 episodes
• On-Policy Distillation — 6× in this paper · explained in 3 episodes
• GRPO — 5× in this paper · explained in 6 episodes
• Knowledge Distillation — 5× in this paper · explained in 11 episodes
• Generalization — 4× in this paper · explained in 13 episodes
• Reinforcement Learning — 4× in this paper · explained in 4 episodes
• Total Variation — 4× in this paper
• Hallucination — 3× in this paper · explained in 3 episodes
• Conference on Computer Vision and Pattern Recognition — 2× in this paper
• Ablation — 1× in this paper · explained in 1 episode
• Ablation Studies — 1× in this paper · explained in 1 episode
• Core — 1× in this paper · explained in 2 episodes
• Exponential Moving Average (EMA) — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Multimodal Reasoning — 1× in this paper · explained in 5 episodes
• On-policy Self-Distillation (OPSD) — 1× in this paper · explained in 2 episodes
• Sparsity — 1× in this paper · explained in 4 episodes
• Training-Free Framework — 1× in this paper · explained in 2 episodes

Sources

• Qwen3-VL Technical Report
• OPD-V: Visual On-Policy Self-Distillation with Modality Balance
• SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learning
• Thinking with Deltas: Incentivizing Reinforcement Learning via Differential Visual Reasoning Policy
• Distilling the Knowledge in a Neural Network
• GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
• VLM-FO1: Bridging the Gap Between High-Level Reasoning and Fine-Grained Perception in VLMs
• Visual-Advantage On-Policy Distillation for Vision-Language Models
• Med-OPD: Improving Medical Vision-Language Models via Evidence-Aware On-Policy Distillation
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning
• Rethinking Selective Knowledge Distillation
• Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
• Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation
• MiMo-VL Technical Report
• VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation
• Thyme: Think Beyond Images
• Self-Distilled Reasoner: On-Policy Self-Distillation for 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