paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达
登录 DINQ
返回新星榜
YW

Yu Wang

ICLR 2026×9AAAI 2026×3IROS 2026×3CVPR 2026×2ICRA 2026×2COLM 2026强化学习与后训练多模态效率与系统推理对齐与偏好
OROpenReview
新星分资深
57
一作贡献70
核心/参与63
方向广度19
合作网络12
2 一作 · 20 篇 · ICLR 2026 / AAAI 2026 / IROS 2026 / CVPR 2026 / ICRA 2026 / COLM 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

20
作者位次 #5/5COLM 2026Poster· Poster Session 4· Imperial Ballroom· #68

NI Sampling++: Optimizing Token Order for Fast Discrete Diffusion Sampling with Reinforcement Learning

Enshu Liu · Theodore Zhengde Zhao · Xuefei Ning · Zinan Lin · Yu Wang

We propose a reinforcement learning framework to optimize token ordering for efficient sampling of diffusion language models.

摘要

Recently, discrete diffusion language models (dLLMs) have attracted increasing attention as an alternative to autoregressive language models, as they naturally allow tokens to be generated in flexible orders and support parallel decoding. However, early heuristic sampling strategies remain inefficient, as they update only a small subset of tokens at each step. NI Sampling addresses this issue by introducing token order optimization for fast sampling, where a lightweight neural indicator determines which tokens to update at each step. Nevertheless, the training objective of NI Sampling is only a proxy for the ultimate goal (i.e., achieving high sampling efficiency without sacrificing generation quality), and its empirical speedup still falls short of the validated upper bound. To bridge this gap, we reformulate token order optimization as a Markov decision process and optimize the neural indicator using reinforcement learning. Experimental results demonstrate that our method achieves further sampling acceleration compared to NI Sampling.

强化学习与后训练效率与系统代码Diffusion Language Model Sampling; Reinforcement Learning;OpenReview
作者位次 #4/4ICRA 2026Oral· Robotics & Automation

A Centerline-Aligned Frenet Graph Framework for Surface-Based Path Planning in Pipeline Environments

Hao Liu · Gang Liu · Chuan Qin · Yu Wang

推理对齐与偏好OpenReview
作者位次 #5/5IROS 2026Oral· Adaptive Allocation of Computational Resources

D3P: Dynamic Denoising Diffusion Policy Via Reinforcement Learning

Shu-ang Yu · Feng Gao · Yi Wu · Chao Yu · Yu Wang

强化学习与后训练Reinforcement LearningDeep Learning in Grasping and ManipulationMachine Learning for Robot ControlOpenReview
作者位次 #4/5IROS 2026Poster· Representation Learning for Robot Manipulation and Locomotion

Motion-Focused Latent Action Enables Cross-Embodiment VLA Training from Human EgoVideos

RUNZE XU · Yiluo Zhang · Tmall Group of Alibaba) · Yu Wang · Jincheng Yu

可解释性多模态社会与人本Representation LearningTransfer LearningDeep Learning in Grasping and ManipulationOpenReview
一作IROS 2026Poster· Vision Systems for Robotics and Industrial Applications

PARKOUR: Semantic-Native Navigation for Underground Parking Via Lightweight OSM and Uncertainty-Aware EKF

Yu Wang · liu xinhao · Xiansheng Guo · huang xia · Haonan Si · Nirwan Ansari

多模态Autonomous Vehicle NavigationLocalizationVision-Based NavigationOpenReview
一作AAAI 2026Poster· Artificial Intelligence

Boosting Noisy Correspondence Discrimination via Dynamic Neighborhood Semantic Verification

Yu Wang · Fengxia Han · Jianyu Wang

OpenReview

常见合作者

CYChao Yu×6XNXuefei Ning×4ZLZinan Lin×4MGMo Guang×4KLKaiwen Long×4ZXZelai Xu×3ELEnshu Liu×2GDGuohao Dai×2SYShu-ang Yu×2HHHaojia Hui×2HYHuazhong Yang×2KHKaiyi Huang×2