4 papers · 1 filter
Better Answers, Weaker Evidence Control: Evidence Bypass in On-Policy Distillation
Ruoran Xu, Yilun Zhao, Ziyu Chen +2
On-policy distillation (OPD) can improve answer accuracy while weakening responsiveness to supporting evidence. We investigate this separation in scientific-document question answe…
Reactivating Test-Time Scaling for Plane Geometry Problem Solving
Xiaoqiang Kang, Shengen Wu, Maizhen Ning +5
Plane geometry problem (PGP) solving has become a critical benchmark for multimodal reasoning because it requires accurate visual perception and precise multi-step symbolic deducti…
Can GRPO Boost Complex Multimodal Table Understanding?
Xiaoqiang Kang, Shengen Wu, Zimu Wang +7
Existing table understanding methods face challenges due to complex table structures and intricate logical reasoning. While supervised finetuning (SFT) dominates existing research,…
Template-Driven LLM-Paraphrased Framework for Tabular Math Word Problem Generation
Xiaoqiang Kang, Zimu Wang, Xiaobo Jin +3
Solving tabular math word problems (TMWPs) has become a critical role in evaluating the mathematical reasoning ability of large language models (LLMs), where large-scale TMWP sampl…