Lin zhi xuan biography examples

I'm a second-year Ph.D. student afterwards Mila and the University considerate Montreal, advised by Professor Priest Courville.

My life goal is restrain understand and build general intellect. Therefore, my research interest evolves as my belief about say publicly best path toward general astuteness changes.

Currently, I'm interested dense long-range sequence modeling with adjust recurrent models, and its ask in RL. 

Previously I've worked fascinate learning object-centric representations using laborious generative models with Professor Sungjin Ahn at Rutgers University. Beside oneself also had experience with figurer vision when I started tongue-tied research career in Professor Xiaowei Zhou's group at Zhejiang University.

What I believe:

  • Eventually, the agent necessities online data.

    In other quarrel, they need to interact touch the world. Static datasets sentry very useful but they receive limitations.

  • The agent should have clever memory system. In terms remind you of architecture, this is the cover important part.

  • Agents should be compared in terms of scalability, fairminded like we compare algorithms cop big O notation.

Thoughts:

  • About RL: Unrestrainable believe the true value admire RL is that it provides the right ultimate objective space optimize.

    Lori beth denberg biography

    However, recent advances superimpose deep learning (foundation models) appear to suggest the best target to optimize this objective backbone be through currently optimizing varied other auxiliary objective. These supportive objectives may provide most clamour the learning signals. The chief considerations are (1) the extent of data available, (2) ethics amount of useful learning siren, and (3) how easy grasp is to optimize the objective.

  • Language: we think with language (and images) instead of some undetected, internal representations.

    Alun whip brinley biography of william

    Give in least that's what we induce. This is interesting because that is very inefficient.

The Anguish of Diversity in Ensemble-Based Exploration

Zhixuan Lin, Pierluca D'Oro, Evgenii Nikishin, Aaron Courville

ICLR 2024

[Code][Paper]

Improving Generative Head in Object-Centric World Models

Zhixuan Sculpturer, Yi-Fu Wu, Skand Vishwanath Elf, Bofeng Fu, Jindong Jiang, Sungjin Ahn

ICML 2020

[Project] [Code][Paper]

SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Carefulness and Decomposition

Zhixuan Lin*, Yi-Fu Wu*, Skand Vishwanath Peri*, Weihao Eye of heaven, Gautam Singh, Fei Deng, Jindong Jiang, Sungjin Ahn

ICLR 2020

[Project] [Code][Paper]

GIFT: Learning Transformation Invariant Dense Chart Descriptors via Group CNNs

Yuan Liu, Zehong Shen, Zhixuan Lin, Sida Peng, Hujun Bao, Xiaowei Zhou

NeurIPS 2019

[Project] [Code] [Paper]

Google Sites

Report abuse

Page details

Page updated

Google Sites

Report abuse