TBNilles/TinyRecursiveModels
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Alexia Jolicoeur-Martineau
2025-10-07 09:30:42 -04:00
committed by GitHub
parent 8120f2bdf7
commit 3302f7de2c
1 changed files with 5 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
README.md
View File

@@ -8,10 +8,6 @@ This is the codebase for the paper: "Less is More: Recursive Reasoning with Tiny
Tiny Recursion Model (TRM) recursively improves its predicted answer y with a tiny network. It starts with the embedded input question x and initial embedded answer y and latent z. For up to K improvements steps, it tries to improve its answer y. It does so by i) recursively updating n times its latent z given the question x, current answer y, and current latent z (recursive reasoning), and then ii) updating its answer y given the current answer y and current latent z. This recursive process allows the model to progressively improve its answer (potentially addressing any errors from its previous answer) in an extremely parameter-efficient manner while minimizing overfitting. Tiny Recursion Model (TRM) recursively improves its predicted answer y with a tiny network. It starts with the embedded input question x and initial embedded answer y and latent z. For up to K improvements steps, it tries to improve its answer y. It does so by i) recursively updating n times its latent z given the question x, current answer y, and current latent z (recursive reasoning), and then ii) updating its answer y given the current answer y and current latent z. This recursive process allows the model to progressively improve its answer (potentially addressing any errors from its previous answer) in an extremely parameter-efficient manner while minimizing overfitting.
<p align="center">
<img src="{{ site.baseurl }}/assets/images/TRM_fig.png" alt="TRM-Figure" style="width:50%">
</p>
### Requirements ### Requirements
- Python 3.10 (or similar) - Python 3.10 (or similar)
@@ -119,14 +115,14 @@ arch.H_cycles=3 arch.L_cycles=4 \
If you find our work useful, please consider citing: If you find our work useful, please consider citing:
```bibtex ```bibtex
@misc{jolicoeurmartineau2025tinyrecursionmodel, @misc{jolicoeurmartineau2025morerecursivereasoningtiny,
title={Less is More: Recursive Reasoning with Tiny Networks}, title={Less is More: Recursive Reasoning with Tiny Networks},
author={Alexia Jolicoeur-Martineau}, author={Alexia Jolicoeur-Martineau},
year={2025}, year={2025},
eprint={xxxxxxx}, eprint={2510.04871},
archivePrefix={arXiv}, archivePrefix={arXiv},
primaryClass={cs.AI}, primaryClass={cs.LG},
url={https://arxiv.org/abs/xxxxxxxxx}, url={https://arxiv.org/abs/2510.04871},
} }
``` ```
@@ -144,4 +140,4 @@ and the Hierarchical Reasoning Model (HRM):
} }
``` ```
This code is based on the Hierarchical Reasoning Model [code](https://github.com/sapientinc/HRM) and the Hierarchical Reasoning Model Analysis [code](https://github.com/arcprize/hierarchical-reasoning-model-analysis). This code is based on the Hierarchical Reasoning Model [code](https://github.com/sapientinc/HRM) and the Hierarchical Reasoning Model Analysis [code](https://github.com/arcprize/hierarchical-reasoning-model-analysis).