AttA2C

KO | EN

gitlite — search

#python #ai #typescript #llm #machine-learning #docker #java #nodejs #kubernetes #vue #pytorch #mcp #openai #open-source

(★ 29)

Attention-based Curiosity-driven Exploration in Deep Reinforcement Learning

AttA2C 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# AttA2C - Attention-based Curiosity-driven Exploration in Deep Reinforcement Learning
#### Author: Patrik Reizinger, MSc student in Electrical Engineering
#### Supervisor and Co-author: Márton Szemenyei, lecturer
#### Organization: Budapest University of Technology and Economics, Department of Control Engineering and Information Technology

Supplementary material for the paper _Attention-based Curiosity-driven Exploration in Deep Reinforcement Learning_ submitted to ICASSP 2020. Preprint available at [https://arxiv.org/abs/1910.10840](https://arxiv.org/abs/1910.10840).

## Table of contents
* [General](#general)
* [Proposed methods](#proposed-methods)
* [Results](#results)
* [Cite](#cite)

## General
The aim of the project is to develop new exploration strategies for Reinforcement Learning for agents which can generalize better. The focus is on curiosity-based methods, such as the Intrinsic Curiosity Module of the paper [Curiosity-driven Exploration by Self-supervised Prediction](https://arxiv.org/abs/1705.05363), which is used extensively to build upon.

This project is implemented in PyTorch, using the [stable-baselines](https://stable-baselines.readthedocs.io/en/master/index.html) package for benchmarking.

## Proposed methods

- __AttA2C (Attention-aided A2C)__: this new version of A2C utilizes attention to split the features fed into the Actor and the Critic.
- __Action- and state-selective ICM__: the extension of ICM aims to use attention for selectively use the features and actions in the forward and inverse dynamic models.
- __Rational Curiosity Module (RCM)__: this novel curiosity formulation aim to incentivize the agent to exploit curiosity only if it contributes to generalization.

## Results
Experiments were carried out on three Atari games: Breakout, Pong and Seaquest (v0 and v4 variants, the former is stochastic, as it uses action repeat with p=0.25).

### Breakout

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/BreakoutNoFrameskip-v0/mean_reward_BreakoutNoFrameskip-v0.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/BreakoutNoFrameskip-v0/feat_std_BreakoutNoFrameskip-v0.png)|
| - | - |

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/BreakoutNoFrameskip-v4/mean_reward_BreakoutNoFrameskip-v4.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/BreakoutNoFrameskip-v4/feat_std_BreakoutNoFrameskip-v4.png)|
| - | - |

### Pong

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/PongNoFrameskip-v0/mean_reward_PongNoFrameskip-v0.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/PongNoFrameskip-v0/feat_std_PongNoFrameskip-v0.png)|
| - | - |

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/PongNoFrameskip-v4/mean_reward_PongNoFrameskip-v4.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/PongNoFrameskip-v4/feat_std_PongNoFrameskip-v4.png)|
| - | - |

### Seaquest

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/SeaquestNoFrameskip-v0/mean_reward_SeaquestNoFrameskip-v0.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/SeaquestNoFrameskip-v0/feat_std_SeaquestNoFrameskip-v0.png)|
| - | - |

|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/SeaquestNoFrameskip-v4/mean_reward_SeaquestNoFrameskip-v4.png)|![](https://raw.githubusercontent.com/rpatrik96/AttA2C/master/figures/SeaquestNoFrameskip-v4/feat_std_SeaquestNoFrameskip-v4.png)|
| - | - |

## Cite
If you found this work useful, please cite the following paper:
```
@article{reizinger2019attention,
  title={Attention-based Curiosity-driven Exploration in Deep Reinforcement Learning},
  author={Reizinger, Patrik and Szemenyei, M{\'a}rton},
  journal={arXiv preprint arXiv:1910.10840},
  year={2019}
}
```

AttA2C

AttA2C 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

AttA2C

AttA2C 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README