7 papers
Improving OCR Faithfulness via Gated and Attenuated On-Policy Distillation
Baode Wang, Zuming Huang, Kexuan Ren +2
Vision-language models may rewrite anomalous text in images into linguistically plausible expressions, compromising OCR transcription faithfulness. Sequence-level task rewards and…
Infinity-Parser2 Technical Report
Zuming Huang, Jun Huang, Kexuan Ren +12
We present Infinity-Parser2, a large multimodal model that couples a controllable data-synthesis pipeline with multi-task reinforcement learning for end-to-end document parsing, ad…
Trust the Right Teacher: Quality-Aware Self-Distillation for GUI Grounding
Jingyuan Huang, Zuming Huang, Yucheng Shi +4
GUI grounding requires vision-language models (VLMs) to identify small target elements in high-resolution screenshots and predict precise screen coordinates. On-policy self-distill…
DyJR: Preserving Diversity in Reinforcement Learning with Verifiable Rewards via Dynamic Jensen-Shannon Replay
Long Li, Zhijian Zhou, Tianyi Wang +7
While Reinforcement Learning (RL) enhances Large Language Model reasoning, on-policy algorithms like GRPO are sample-inefficient as they discard past rollouts. Existing experience…
Infinity Parser: Layout Aware Reinforcement Learning for Scanned Document Parsing
Baode Wang, Biao Wu, Weizhen Li +8
Document parsing from scanned images into structured formats remains a significant challenge due to its complexly intertwined elements such as text paragraphs, figures, formulas, a…
Infinity Parser: Layout Aware Reinforcement Learning for Scanned Document Parsing
Baode Wang, Biao Wu, Weizhen Li +8
Automated parsing of scanned documents into richly structured, machine-readable formats remains a critical bottleneck in Document AI, as traditional multi-stage pipelines suffer fr…
VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning
Haozhe Wang, Chao Qu, Zuming Huang +3
Recently, slow-thinking systems like GPT-o1 and DeepSeek-R1 have demonstrated great potential in solving challenging problems through explicit reflection. They significantly outper…