paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达

更新于

返回新星榜

MS

Masashi Sugiyama

ICML 2026×7强化学习与后训练社会与人本推理对齐与偏好理论与学习动力学

新星分资深

26

一作贡献0

核心/参与10

方向广度10

合作网络12

0 一作 · 7 篇 · ICML 2026

获取完整研究档案

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。

联系方式

实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。

2026 顶会论文

7

作者位次 #6/6ICML 2026Poster· Reinforcement Learning / Online

Reinforcement Learning from Bagged Reward

Yuting Tang · Xin-Qiang Cai · Yao-Xiang Ding · Qiyu Wu · Guoqing Liu · Masashi Sugiyama

强化学习与后训练OpenReview

作者位次 #6/6ICML 2026Poster· Deep Learning / Large Language Models

Decomposing the Basic Abilities of Large Language Models: Mitigating Cross-Task Interference in Multi-Task Instruct-Tuning

Bing Wang · Ximing Li · Changchun Li · Jinjin Chi · Gang Niu · Masashi Sugiyama

作者位次 #4/4ICML 2026Poster· Deep Learning / Large Language Models

Gradient Regularization Mitigates Reward Hacking in Reinforcement Learning from Human Feedback and Verifiable Rewards

Johannes Ackermann · Michael Noukhovitch · Takashi Ishida · Masashi Sugiyama

强化学习与后训练社会与人本OpenReview

作者位次 #6/6ICML 2026Poster

Positive-Unlabeled Learning with Extreme Scarcity of Labeled Positives

Yuanchao Dai · Ximing Li · Wei Wang · Changchun Li · Gang Niu · Masashi Sugiyama

作者位次 #10/11ICML 2026Poster· Reinforcement Learning / Everything Else

Positive–Unlabeled Reinforcement Learning Distillation for On-Premise Small Models

Zhiqiang Kou · Junyang Chen · Xin-Qiang Cai · Xiaobo Xia · Ming-Kun Xie · Dong-Dong Wu · Biao Liu · Yuheng Jia · Xin Geng · Masashi Sugiyama · Tat-Seng Chua

强化学习与后训练OpenReview

作者位次 #27/29ICML 2026Poster· Deep Learning / Foundation Models

Proteo-R1: Reasoning Foundation Models for De Novo Protein Design

Fang Wu · Weihao Xuan · Heli Qi · Hanqun CAO · Heng-Jui Chang · Zeqi Zhou · Li Erran Li · Haokai Zhao · Jian Ma · Zijian Ma · Yu-Chi Cheng · Kuan Pang · Robert Tang · Zehong Wang · Guanlue Li · Hanchen Wang · Kejun Ying · Pan Lu · Chiho Im · Seungju Han · Peng Xia · Tinson Xu · Yinxi Li · Deyao Zhu · Pheng Ann Heng · Naoto Yokoya · Masashi Sugiyama · Jure Leskovec · Yejin Choi

推理OpenReview

作者位次 #4/4ICML 2026Poster· Theory / Domain Adaptation and Transfer Learning

Incorporating Importance Weighting in Optimal Transport Based Domain Alignment

Okan Koç · Alexander Soen · Shanglin Li · Masashi Sugiyama

对齐与偏好理论与学习动力学OpenReview

常见合作者

XCXin-Qiang Cai×2 XLXiming Li×2 CLChangchun Li×2 GNGang Niu×2 YTYuting Tang YDYao-Xiang Ding QWQiyu Wu GLGuoqing Liu BWBing Wang JCJinjin Chi JAJohannes Ackermann MNMichael Noukhovitch