paperlist.ai

NeurIPS 2020 论文列表

Conference on Neural Information Processing Systems

202620252024202320222021202020192018201720162015

录用类型

1,901篇论文

按官方录用论文列表统计，可能与会议公布的数字略有出入。

  • Oral
    1065.6%
  • Spotlight
    28014.7%
  • Poster
    1,51379.6%
  • 其他 / 未标注
    20.1%

Top 10 主题（点击筛选）

全部 207 个研究方向的趋势

浏览全部论文

全部1,901Oral106Spotlight280Poster1,513

主题: 世界模型 · 2 篇论文 · 重置筛选

Latent World Models For Intrinsically Motivated Exploration

Aleksandr Ermolov, Nicu Sebe

展开摘要

In this work we consider partially observable environments with sparse rewards. We present a self-supervised representation learning method for image-based observations, which arranges embeddings respecting temporal distance of observations. This representation is empirically robust to stochasticity and suitable for novelty detection from the error of a predictive forward model. We consider episodic and life-long uncertainties to guide the exploration. We propose to estimate the missing information about the environment with the world model, which operates in the learned latent space. As a motivation of the method, we analyse the exploration problem in a tabular Partially Observable Labyrinth. We demonstrate the method on image-based hard exploration environments from the Atari benchmark and report significant improvement with respect to prior work. The source code of the method and all the experiments is available at https://github.com/htdt/lwm.

Spotlight
世界模型
Bridging Imagination and Reality for Model-Based Deep Reinforcement Learning

Guangxiang Zhu, Minghao Zhang, Honglak Lee, Chongjie Zhang

展开摘要

Sample efficiency has been one of the major challenges for deep reinforcement learning. Recently, model-based reinforcement learning has been proposed to address this challenge by performing planning on imaginary trajectories with a learned world model. However, world model learning may suffer from overfitting to training trajectories, and thus model-based value estimation and policy search will be prone to be sucked in an inferior local policy. In this paper, we propose a novel model-based reinforcement learning algorithm, called BrIdging Reality and Dream (BIRD). It maximizes the mutual information between imaginary and real trajectories so that the policy improvement learned from imaginary trajectories can be easily generalized to real trajectories. We demonstrate that our approach improves sample efficiency of model-based planning, and achieves state-of-the-art performance on challenging visual control benchmarks.

Poster
世界模型