BAR: A Backward Reasoning based Agent for Complex Minecraft Tasks

Our proposed backward reasoning based agent

About BAR

Large language model (LLM) based agents have shown great potential in following human instructions and automatically completing various tasks. To complete a task, the agent needs to decompose it into easily executed steps by planning. Existing studies mainly conduct the planning by inferring what steps should be executed next starting from the agent's initial state. However, this forward reasoning paradigm doesn't work well for complex tasks. We propose to study this issue in Minecraft, a virtual environment that simulates complex tasks based on real-world scenarios. We believe that the failure of forward reasoning is caused by the big perception gap between the agent's initial state and task goal. To this end, we leverage backward reasoning and make the planning starting from the terminal state, which can directly achieve the task goal in one step. Specifically, we design a BAckward Reasoning based agent (BAR). It is equipped with a recursive goal decomposition module, a state consistency maintaining module and a stage memory module to make robust, consistent, and efficient planning starting from the terminal state.

Install Dependencies

Prepare the Environment

We recommend using Anaconda to manage the environment. If you don't have Anaconda installed, you can download it from here.

conda create -n bar python=3.10
conda activate bar

Make sure you have JDK 8 installed. If you don't have it installed, you can install it using the following command:

conda install openjdk=8

To check your JDK version, run the command java -version:

openjdk version "1.8.0_392"
OpenJDK Runtime Environment (build 1.8.0_392-8u392-ga-1~20.04-b08)
OpenJDK 64-Bit Server VM (build 25.392-b08, mixed mode)

Install BAR

Clone the repository:

git clone https://github.com/WitcherLeo/BAR.git

As we use the same environment as JARVIS-1, you need to download the MCP environment from this link
Then unzip the file stark_tech.zip to the path jarvis/stark_tech

Then you can install BAR as a Python package.

pip install -e .

Install the dependencies:

pip install -r requirements.txt

Usage

You need to set the environment variable OPENAI_API_KEY and OPENAI_API_BASE first.

export OPENAI_API_KEY="sk-******"
export OPENAI_API_BASE="your openai base url"

Then you can run the following command to start the static planning for the BAR agent.

python -m bar.run_planner --llm_name GPT4 --cuda_ip 0 --groups stone iron diamond --overwrite

To-Do

• Release dynamic planning environment and code.
• Release customization BAR to enable users to customize their own BAR agents.

Check out our paper!

Our paper is available on Arxiv. Please cite our paper if you find BAR useful for your research:

@article{du2025bar,
  title={BAR: A Backward Reasoning based Agent for Complex Minecraft Tasks},
  author={Du, Weihong and Liao, Wenrui and Yan, Binyu and Liang, Hongru and Cohn, Anthony G and Lei, Wenqiang},
  journal={arXiv preprint arXiv:2505.14079},
  year={2025}
}