作者位次 #3/5ICML 2026Poster· Deep Learning / Large Language Models
ConPress: Learning Efficient Reasoning from Multi-Question Contextual Pressure
Jie Deng · Shining Liang · Jun Li · Hongzhi Li · Yutao Xie
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Jie Deng · Shining Liang · Jun Li · Hongzhi Li · Yutao Xie
Jie Deng · Hanshuang Tong · Jun Li · Shining Liang · Ning Wu · Hongzhi Li · Yutao Xie
Large language models (LLMs) have made impressive strides in mathematical reasoning, often fine-tuned using rejection sampling, which retains only correct reasoning trajectories. While effective, this paradigm treats supervision as a binary filter that systematically excludes teacher-generated er...