Researcher — paper.dinq.me

2026 顶会论文

作者位次 #5/5ICML 2026Oral

CoEvol-NO: State and Coordinate Co-Evolution with an Error-Driven Predictor-Corrector Paradigm for Neural Operator Transformer

Jianqiao Zeng · Ruocheng Wang · Yanzhi Liu · Hao Xiong · Junchi Yan

OpenReview

作者位次 #3/3ICML 2026Poster· Applications / Chemistry, Physics, and Earth Sciences

MOD-SR: Unifying Multimodal Learning and Direct Optimization with Gradient-Guided Diffusion Model for Symbolic Regression

Chuyang Xiang · Yichen Wei · Junchi Yan

多模态OpenReview

作者位次 #3/3ICML 2026Poster· Deep Learning / Algorithms

IO-Adam: Rethinking Memory-Efficient Adaptive Optimizers from Gradient Computation

Yiting Chen · Zongwei Huo · Junchi Yan

效率与系统检索与知识OpenReview

作者位次 #6/6ICML 2026Poster· Applications / Chemistry, Physics, and Earth Sciences

G-RANS: Generalizable Residual-Aware Neural Solvers for Sparse Systems

Weixin Liao · Mingquan Feng · Zhizhou Zhang · Youjia Wu · Yifan Fu · Junchi Yan

理论与学习动力学OpenReview

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

NITP: Next Implicit Token Prediction for LLM Pre-training

Xiangdong Zhang · Debing Zhang · Shaofeng Zhang · Xiaohan Qin · Yu Cheng · Junchi Yan

数据与预训练OpenReview

作者位次 #12/12ICML 2026Poster· Deep Learning / Large Language Models

Attention Illuminates LLM Reasoning: The Uncovered Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization

Yang Li · Zhichen Dong · Yuhan Sun · Weixun Wang · Shaopan Xiong · Yijia Luo · Jiashun Liu · Han Lu · Jiamang Wang · Wenbo Su · Bo Zheng · Junchi Yan

推理强化学习与后训练OpenReview

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

Beyond Theorem Proving: Formulation, Framework and Benchmark for Formal Problem-Solving

Qi Liu · Xinhao Zheng · Renqiu Xia · Xingzhi Qi · Qinxiang Cao · Junchi Yan

评测与基准理论与学习动力学OpenReview

作者位次 #5/5ICML 2026Poster· Deep Learning / Generative Models and Autoencoders

Primal-Spectral Generative Modeling: Fast Analytical Generation via Pseudoinverse Lévy Inversion

Zhiyuan Ouyang · Simei Huang · Zuokai Wen · Xiangyun Zhang · Junchi Yan

OpenReview

作者位次 #6/6ICML 2026Poster· Applications / Everything Else

Learning to Decode Against Compositional Hallucination in Video Multimodal Large Language Models

Wenbin Xing · Quanxing Zha · Lizheng Zu · Mengran Li · Ming Li · Junchi Yan

多模态安全长上下文与架构OpenReview

作者位次 #7/9ICML 2026Poster

More Sail than Ballast: Addressing Harmful Knowledge Leakage in the Expansive Reasoning Space of LRMs

Qibing Ren · Xinhao Song · Ke Fan · Lijun Li · Zhanpeng Zhou · Gongshen Liu · Junchi Yan · Lizhuang Ma · Jing Shao

推理安全检索与知识OpenReview

作者位次 #3/3ICML 2026Poster· Applications / Everything Else

RulePlanner: All-in-One Reinforcement Learner for Unifying Design Rules in 3D Floorplanning

Ruizhe Zhong · Xingbo Du · Junchi Yan

推理强化学习与后训练OpenReview

作者位次 #5/7ICML 2026Poster· Optimization / Non-Convex

Diffusion-based learning framework for Constrained Nonconvex Optimization with Weighted Bootstrapped Refinement

Shutong Ding · Yimiao Zhou · Ke Hu · Xi Yao · Junchi Yan · Xiaoying Tang · Ye Shi

OpenReview

作者位次 #10/10ICML 2026Poster· Deep Learning / Large Language Models

How Does Reasoning Flow? Tracing Attention-Induced Information Flow for Targeted RL in LLMs

Zhichen Dong · Yang Li · Yuhan Sun · Weixun Wang · Yijia Luo · Zinian Peng · Wenbo Su · YuCheng · Bo Zheng · Junchi Yan

推理强化学习与后训练OpenReview

作者位次 #8/8ICML 2026Poster· Deep Learning / Large Language Models

TodoEvolve: Learning to Architect Agent Planning Systems

Jiaxi Liu · Guibin Zhang · Yanzuo Jiang · Zihan Zhang · Heng Chang · Zhenfei Yin · Qibing Ren · Junchi Yan

推理智能体与工具OpenReview

作者位次 #7/8ICML 2026Poster· Applications / Everything Else

MALICE: Memory-aware Loop Invariants Generation on Symbolic Execution Traces

Tong Chen · Siyu Liu · Hongyi Zhong · liao zhang · Lixiang Wang · Xiwei Wu · Junchi Yan · Qinxiang Cao

检索与知识OpenReview

一作ICML 2026Poster· Optimization / Convex

Design Linear Constrained Neural Layers with Implicit Convex Optimization

Junchi Yan · Jiaxi Liu · Yihui Tu · Fangyuan Zhou · Wenzheng Pan · Zhongteng Gui · Liangliang Shi

OpenReview

作者位次 #5/5ICML 2026Poster· Optimization / Discrete and Combinatorial Optimization

Problem Distributions as Tasks: Repurposing Meta Learning for Generative Combinatorial Optimization towards Multi-task Pretraining and Adaptation

Wenzheng Pan · Jiale Ma · Nuoyan Chen · Yang Li · Junchi Yan

数据与预训练OpenReview

作者位次 #8/8ICML 2026Poster· Deep Learning / Large Language Models

Inner-layer Token Self-modulation as Another Scaling Axis for LLMs

Yebin Yang · Huaijin Wu · Jingtao Han · Yu Wang · Xiaohan Qin · Jingzhi Wang · Debing Zhang · Junchi Yan

OpenReview

作者位次 #6/10ICML 2026Poster· Deep Learning / Large Language Models

Towards Cold-Start Drafting and Continual Refining: A Value-Driven Memory Approach with Application to NPU Kernel Synthesis

Yujie Zheng · Zhuo Li · Shengtao Zhang · Jiaqian Wang · Junjie Sheng · Junchi Yan · Weinan Zhang · Ying Wen · Bo Tang · Muning Wen

检索与知识OpenReview

作者位次 #9/9CVPR 2026Poster· Computer Vision & Pattern Recognition

SoliReward: Mitigating Susceptibility to Reward Hacking and Annotation Noise in Video Generation Reward Models

Jiesong Lian · Ruizhe Zhong · Zixiang Zhou · Xiaoyue Mi · Long Hu · Yuan Zhou · Qinglin Lu · Yixue Hao · Junchi Yan

多模态对齐与偏好OpenReview 项目页

Junchi Yan

获取完整研究档案

联系方式

2026 顶会论文

CoEvol-NO: State and Coordinate Co-Evolution with an Error-Driven Predictor-Corrector Paradigm for Neural Operator Transformer

MOD-SR: Unifying Multimodal Learning and Direct Optimization with Gradient-Guided Diffusion Model for Symbolic Regression

IO-Adam: Rethinking Memory-Efficient Adaptive Optimizers from Gradient Computation

G-RANS: Generalizable Residual-Aware Neural Solvers for Sparse Systems

NITP: Next Implicit Token Prediction for LLM Pre-training

Attention Illuminates LLM Reasoning: The Uncovered Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization

Beyond Theorem Proving: Formulation, Framework and Benchmark for Formal Problem-Solving

Primal-Spectral Generative Modeling: Fast Analytical Generation via Pseudoinverse Lévy Inversion

Learning to Decode Against Compositional Hallucination in Video Multimodal Large Language Models

More Sail than Ballast: Addressing Harmful Knowledge Leakage in the Expansive Reasoning Space of LRMs

RulePlanner: All-in-One Reinforcement Learner for Unifying Design Rules in 3D Floorplanning

Diffusion-based learning framework for Constrained Nonconvex Optimization with Weighted Bootstrapped Refinement

How Does Reasoning Flow? Tracing Attention-Induced Information Flow for Targeted RL in LLMs

TodoEvolve: Learning to Architect Agent Planning Systems

MALICE: Memory-aware Loop Invariants Generation on Symbolic Execution Traces

Design Linear Constrained Neural Layers with Implicit Convex Optimization

Problem Distributions as Tasks: Repurposing Meta Learning for Generative Combinatorial Optimization towards Multi-task Pretraining and Adaptation

Inner-layer Token Self-modulation as Another Scaling Axis for LLMs

Towards Cold-Start Drafting and Continual Refining: A Value-Driven Memory Approach with Application to NPU Kernel Synthesis

SoliReward: Mitigating Susceptibility to Reward Hacking and Annotation Noise in Video Generation Reward Models

常见合作者