Infinity-Parser: Layout-Aware Reinforcement Learning with High-quality Document Parsing Dataset (2026 · 4 citations)
Converting scanned pages into structured Markdown is difficult. Pages mix paragraphs, tables, formulas and figures, and a parser must recover both the content and its layout and reading order. The authors identify two obstacles: there is little training data with precise, layout-aware annotation, and supervised…
Mock Worlds, Real Skills: Building Small Agentic Language Models with Synthetic Tasks, Simulated Environments, and Rubric-Based Rewards (2026 · 1 citation)
Small language models usually lag well behind large, expensive LLMs on agentic tool-use tasks. Reinforcement learning could help, but two obstacles hold it back. Public agentic datasets are narrow and often too easy, so RL rollouts give little learning signal. Real APIs are also costly, unstable and limited in scope.…
AgenticQwen: Training Small Agentic Language Models with Dual Data Flywheels for Industrial-Scale Tool Use (2026)
Industrial agent systems often run on large proprietary or open models, which costs too much at scale. Major labs rarely release small models with strong tool-use skills. This Alibaba paper presents AgenticQwen, a family of compact agents that use small Qwen models as their foundation. Their training relies on…

Co-authors here

12 co-authors with a page, the most papers together first.

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

Papers of our library cite their works 1 time (of the 149,439 papers whose bibliographies we read); 1 citing paper with a page here →.

Updated 1 November 2026