Montezuma's Revenge

28 papers with code • 1 benchmarks • 1 datasets

Montezuma's Revenge is an ATARI 2600 Benchmark game that is known to be difficult to perform on for reinforcement learning algorithms. Solutions typically employ algorithms that incentivise environment exploration in different ways.

For the state-of-the art tables, please consult the parent Atari Games task.

( Image credit: Q-map )

Most implemented papers

Rainbow: Combining Improvements in Deep Reinforcement Learning

thu-ml/tianshou 6 Oct 2017

The deep reinforcement learning community has made several independent improvements to the DQN algorithm.

Exploration by Random Network Distillation

openai/random-network-distillation ICLR 2019

In particular we establish state of the art performance on Montezuma's Revenge, a game famously difficult for deep reinforcement learning methods.

Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation

transedward/pytorch-hdqn NeurIPS 2016

Learning goal-directed behavior in environments with sparse feedback is a major challenge for reinforcement learning algorithms.

Go-Explore: a New Approach for Hard-Exploration Problems

uber-research/go-explore 30 Jan 2019

Go-Explore can also harness human-provided domain knowledge and, when augmented with it, scores a mean of over 650k points on Montezuma's Revenge.

Scaling All-Goals Updates in Reinforcement Learning Using Convolutional Neural Networks

fabiopardo/qmap ICLR 2019

Being able to reach any desired location in the environment can be a valuable asset for an agent.

Exploring Unknown States with Action Balance

NeteaseFuxiRL/action-balance-exploration 10 Mar 2020

In this paper, we focus on improving the effectiveness of finding unknown states and propose action balance exploration, which balances the frequency of selecting each action at a given state and can be treated as an extension of upper confidence bound (UCB) to deep reinforcement learning.

First return, then explore

uber-research/go-explore 27 Apr 2020

The promise of reinforcement learning is to solve complex sequential decision problems autonomously by specifying a high-level reward function only.

Reinforcement Learning with Latent Flow

WendyShang/flare NeurIPS 2021

Temporal information is essential to learning effective policies with Reinforcement Learning (RL).

A Study of Global and Episodic Bonuses for Exploration in Contextual MDPs

facebookresearch/e3b 5 Jun 2023

This results in an algorithm which sets a new state of the art across 16 tasks from the MiniHack suite used in prior work, and also performs robustly on Habitat and Montezuma's Revenge.

Unifying Count-Based Exploration and Intrinsic Motivation

RLAgent/state-marginal-matching NeurIPS 2016

We consider an agent's uncertainty about its environment and the problem of generalizing this uncertainty across observations.