Fewer Tokens, More Self-Teaching: On-Policy Self-Distillation for Extreme Visual Token Reduction

Junxian Li; Ruixuan Yang; Tianao Zhang; Tiange Xu; Weisheng Dong; Yulun Zhang

title	Fewer Tokens, More Self-Teaching: On-Policy Self-Distillation for Extreme Visual Token Reduction
authors	Junxian Li, Ruixuan Yang, Tianao Zhang, Tiange Xu, Weisheng Dong, Yulun Zhang
date	2026-09-26
venue	arXiv (preprint, not peer reviewed)
id	arXiv:2609.32353
affiliation	Shanghai Jiao Tong University
upvotes	8 (Hugging Face Daily Papers, 2026-11-01)
links	abs pdf hf code
abstract	Visual token reduction is an effective way to accelerate multimodal large language models (MLLMs), but performance deteriorates rapidly under extremely low token budgets. Existing work has explored both visual-token selection and training-based adaptation to reduced visual inputs. We take a step further by asking how a heavily compressed MLLM should learn from the states induced by its own generations. This setting naturally calls for on-policy self-distillation: a heavily compressed model is supervised on the states induced by its own generations, while its full-token counterpart serves as an information-rich teacher. Based on this insight, we propose LT-OPD, a training framework for extreme visual-token reduction. The student rolls out responses with only a small fraction of visual tokens, and a frozen full-token copy of the same MLLM provides distributional supervision along these student-generated trajectories. To stabilize on-policy learning when visual evidence is severely limited, we further introduce a budget-level curriculum that progressively decreases the token budget during training. Across nine benchmarks on Qwen3.5-4B, LT-OPD raises average retained performance under 5% visual-token retention from 68.6% to 82.3%, outperforming training-free, training-based, and reinforcement-learning baselines at the same budget. The gains transfer consistently to Qwen3.5-9B, GLM-4.6V-9B, and LLaVA-OV-1.5-4B. LT-OPD also reduces KV-cache usage by 85.2% and prefill FLOPs by 85.4% without additional inference overhead, demonstrating that on-policy learning can substantially recover capabilities lost to extreme visual-token reduction.

cite

@misc{li2026fewer,
  title={Fewer Tokens, More Self-Teaching: On-Policy Self-Distillation for Extreme Visual Token Reduction},
  author={Junxian Li and Ruixuan Yang and Tianao Zhang and Tiange Xu and Weisheng Dong and Yulun Zhang},
  year={2026},
  eprint={2609.32353},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2609.32353}
}