Beyond Surprise: Improving Exploration Through Surprise Novelty

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Le, Hung, Do, Kien, Nguyen, Dung, Venkatesh, Svetha
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909088989839360
author Le, Hung
Do, Kien
Nguyen, Dung
Venkatesh, Svetha
author_facet Le, Hung
Do, Kien
Nguyen, Dung
Venkatesh, Svetha
contents We present a new computing model for intrinsic rewards in reinforcement learning that addresses the limitations of existing surprise-driven explorations. The reward is the novelty of the surprise rather than the surprise norm. We estimate the surprise novelty as retrieval errors of a memory network wherein the memory stores and reconstructs surprises. Our surprise memory (SM) augments the capability of surprise-based intrinsic motivators, maintaining the agent's interest in exciting exploration while reducing unwanted attraction to unpredictable or noisy observations. Our experiments demonstrate that the SM combined with various surprise predictors exhibits efficient exploring behaviors and significantly boosts the final performance in sparse reward environments, including Noisy-TV, navigation and challenging Atari games.
format Preprint
id arxiv_https___arxiv_org_abs_2308_04836
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Beyond Surprise: Improving Exploration Through Surprise Novelty
Le, Hung
Do, Kien
Nguyen, Dung
Venkatesh, Svetha
Machine Learning
We present a new computing model for intrinsic rewards in reinforcement learning that addresses the limitations of existing surprise-driven explorations. The reward is the novelty of the surprise rather than the surprise norm. We estimate the surprise novelty as retrieval errors of a memory network wherein the memory stores and reconstructs surprises. Our surprise memory (SM) augments the capability of surprise-based intrinsic motivators, maintaining the agent's interest in exciting exploration while reducing unwanted attraction to unpredictable or noisy observations. Our experiments demonstrate that the SM combined with various surprise predictors exhibits efficient exploring behaviors and significantly boosts the final performance in sparse reward environments, including Noisy-TV, navigation and challenging Atari games.
title Beyond Surprise: Improving Exploration Through Surprise Novelty
topic Machine Learning
url https://arxiv.org/abs/2308.04836