# Teach with Blinks: Breaking Linguistic Momentum in On-Policy Distillation for Fine-Grained Multimodal Perception

**Authors:** Tangzheng Lian, Hang Zhang, Guanyu Hu, Renjiu Hu, Zhensong Zhang, Jiankang Deng, Jifei Song  
**Published:** 12 Nov 2026  
**Links:** [arXiv](https://arxiv.org/abs/2611.16402) · [PDF](https://arxiv.org/pdf/2611.16402) · [Code](https://github.com/Supltz/BlinkOPD) · [This page](https://omen-bit.github.io/ai-paper-daily/paper/2611.16402.html)

## Key numbers

- **12×**: …achieves state-of-the-art average performance across six fine-grained benchmarks using only 500 training images (12x less data)
- **500**: BlinkOPD achieves state-of-the-art average performance across six fine-grained benchmarks using only 500 training images

## Abstract

On-policy distillation (OPD) has emerged as an effective post-training approach to improving fine-grained perception in multimodal large language models by amortizing costly inference-time zooming operations into a single-pass recipe. However, vanilla OPD is inherently susceptible to linguistic momentum: the teacher can follow the student's erroneous prefixes, thus weakening corrective supervision. To address this, we propose Blink On-Policy Distillation (BlinkOPD), in which the teacher "blinks" at every token: it reads the same prefix once with its eyes open and once with its eyes closed. A first-glance gain further strengthens the correction on rollouts that conflict with the teacher. By extracting more effective supervision from each on-policy trajectory, BlinkOPD achieves state-of-the-art average performance across six fine-grained benchmarks using only 500 training images (12x less data) than existing OPD methods. Beyond fine-grained perception, BlinkOPD also demonstrates strong zero-shot improvements across ten held-out benchmarks spanning general visual perception, visual hallucination, and multimodal reasoning abilities. Code: https://github.com/Supltz/BlinkOPD

---

Notes written by AI from the paper, numbers copied from the abstract. Check the paper before you rely on them. From AI Papers Daily.