Datasets:

Kalso42
/

WorldModelForMaze

license: mit
task_categories:
  - other
tags:
  - maze
  - world-model
  - sequence-modeling
  - transformer
  - mamba
pretty_name: World Model For Maze

WorldModelForMaze

Code, datasets, and trained checkpoints for studying world-model representations in maze navigation, based on a modified NanoGPT.

*.py — training, testing, probing, and visualization scripts (see readme.md).
model/ — architectures: transformer, transformer-rope, transformer-nextlat, mamba, mamba2, gated-deltanet, gru.
data/maze/100/ — tokenized maze datasets for Tasks A/C/E/H/I (RWs paths, 100 nodes).
out/ — final (10000-iter) checkpoints, maze_kdetour results, and plots.

Intermediate (non-10000-iter) checkpoints (out2/) are not included here.

Usage

See readme.md for full instructions on data generation, training, testing, probing, and the k-step detour analysis.

Quick start:

git clone https://huggingface.co/datasets/Kalso42/WorldModelForMaze
cd WorldModelForMaze
python train_maze.py --tasks C1 --path_type RWs --num_train_dataset 10M --config 6_6_384

WorldModelForMaze

Contents

Usage