# samsungsailmontreal/tinyrecursivemodels

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/samsungsailmontreal-tinyrecursivemodels).**

6,540 stars · 1,033 forks · Python · MIT · archived

## Links

- GitHub: https://github.com/SamsungSAILMontreal/TinyRecursiveModels
- awesome-repositories: https://awesome-repositories.com/repository/samsungsailmontreal-tinyrecursivemodels.md

## Description

TinyRecursiveModels is a recursive training framework for small neural networks designed to solve complex logical tasks. It functions as a parameter-efficient model trainer and a reasoning dataset generator, enabling the optimization of models that refine their answers through iterative reasoning steps.

The framework differentiates itself by utilizing latent-state recursive refinement, where the model maintains and updates an internal hidden representation to improve prediction accuracy over multiple sequential steps. It also includes tools for generating structured training and evaluation datasets based on logical puzzles and maze solving.

The system covers hardware-accelerated training loops and parameter-efficient network design to reduce computational overhead while maintaining reasoning capabilities.

## Tags

### Artificial Intelligence & ML

- [Latent State Recursive Refiners](https://awesome-repositories.com/f/artificial-intelligence-ml/latent-state-recursive-refiners.md) — Implements a recursive framework where the model iteratively updates its internal latent state to refine reasoning accuracy.
- [Latent State Refinement](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-ai-resources/diffusion-visual-models/generative-ai-models/latent-space-generative-models/latent-space-projections/latent-space-encoders/latent-space-manipulations/latent-state-refinement.md) — Improves model predictions by updating an internal latent state across multiple steps before final output.
- [State Trackers](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-ai-resources/diffusion-visual-models/generative-ai-models/latent-space-generative-models/latent-space-projections/latent-space-encoders/state-trackers.md) — Maintains a persistent hidden representation that evolves as the model progresses through iterative reasoning steps.
- [Iterative Prediction Refiners](https://awesome-repositories.com/f/artificial-intelligence-ml/model-predictions/prediction-engines/iterative-prediction-refiners.md) — Provides a mechanism to refine model outputs by iteratively feeding predictions back into the network for successive corrections.
- [Reasoning Model Training Suites](https://awesome-repositories.com/f/artificial-intelligence-ml/reasoning-model-training-suites.md) — Provides a training suite for optimizing small neural networks to solve complex logical puzzles recursively.
- [Synthetic Reasoning Data Generators](https://awesome-repositories.com/f/artificial-intelligence-ml/synthetic-data-generators/synthetic-reasoning-data-generators.md) — Generates structured training data by mapping logical puzzles and maze solutions into reasoning sequences.
- [Neural Network Training Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/neural-network-training-frameworks.md) — Provides a framework for building and optimizing small-scale neural networks on complex logical tasks. ([source](https://cdn.jsdelivr.net/gh/samsungsailmontreal/tinyrecursivemodels@main/README.md))
- [Hardware Training Acceleration](https://awesome-repositories.com/f/artificial-intelligence-ml/neural-network-training/hardware-training-acceleration.md) — Implements hardware-level optimizations for the neural network training process using GPU and TPU acceleration.
- [Parameter-Efficient Training Toolkits](https://awesome-repositories.com/f/artificial-intelligence-ml/parameter-efficient-training-toolkits.md) — Ships a trainer for optimizing small-scale models on difficult tasks using parameter-efficient techniques.
- [Parameter-Efficient Tuning Techniques](https://awesome-repositories.com/f/artificial-intelligence-ml/parameter-efficient-tuning-techniques.md) — Optimizes small-scale models using parameter-efficient techniques to reduce resource requirements.

### Part of an Awesome List

- [Efficient Neural Architectures](https://awesome-repositories.com/f/awesome-lists/ai/efficient-neural-architectures.md) — Employs small-scale, parameter-efficient neural network architectures designed to minimize computational overhead.
- [Recursive Answer Refiners](https://awesome-repositories.com/f/awesome-lists/ai/question-answering-models/recursive-answer-refiners.md) — Improves predicted outputs by iteratively updating internal states and answers through multiple reasoning steps. ([source](https://cdn.jsdelivr.net/gh/samsungsailmontreal/tinyrecursivemodels@main/README.md))
- [Reasoning Dataset Builders](https://awesome-repositories.com/f/awesome-lists/ai/reasoning-datasets/reasoning-dataset-builders.md) — Constructs structured training and evaluation datasets for logical puzzles and maze solving. ([source](https://cdn.jsdelivr.net/gh/samsungsailmontreal/tinyrecursivemodels@main/README.md))