Reframe: Observation-Aligned Multi-Teacher On-Policy Distillation for Agentic Visual Reasoning
cs.CV
Submitted: 2026-11-09
Updated: 2026-11-09
Code: https://github.com/EliYuan30/Reframe
Terminology
Sources
- D cubed-MOPD: Dynamic Domain ScheDuling for Efficient Multi-Teacher Distillation
- DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
- DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
- Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding
- Rethinking On-Policy Distillation of Large Language Models II: One Training Example
- Kimi K3: Open Frontier Intelligence
- Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
- UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI Agents
- PerceptionBench: Evaluating Atomic Visual Perception in Multimodal Large Language Models
- MiMo-V2-Flash Technical Report
- MOPD: Multi-Teacher On-Policy Distillation for Capability Integration in LLM Post-Training
- When Should a VLM Look? Paying Only for Visual Calls That Were Needed and Used
- Video-CoM: Interactive Video Reasoning via Chain of Manipulations
- When Top-K Misses the Decision: Tool-Call Drift in Multi-Teacher On-Policy Distillation
- A Survey of On-Policy Distillation for Large Language Models
- Beacon: Knowing When and How to Perform Agentic Visual Reasoning
- Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation
- STEP-OPD: Rethinking Output Targets and Internal Dynamics in On-Policy Distillation for Diffusion Models
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs