Downloads · 30 days
0
fprm
fixed-point-reasoners/fprm
fprm is a machine learning model from fixed-point-reasoners. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Checkpoints for FPRM / FPTRM (Fixed-Point Tiny Recursive Model), a weight-tied iterative reasoner trained with a fixed-point solver and test-time compute scaling.
Access
Public
Updated Jun 22, 2026
Repo size
47.5 GB
Likes
2
Public
Repo makeup
Click a slice to open those files.
Other44.2 GB · 93%
At a glance
- License
- mit
- Access
- Public
- Created
- Jun 22, 2026
- Updated
- Jun 22, 2026
- SHA
- 0f20ef7c
Model card
From the Hugging Face model README
FPRM — Fixed-Point Tiny Recursive Models
Checkpoints for FPRM / FPTRM (Fixed-Point Tiny Recursive Model), a weight-tied iterative reasoner trained with a fixed-point solver and test-time compute scaling.
Each subfolder is one self-contained run (checkpoints + exact config + model source +
reproduction scripts). See each folder's README.md for the full recipe and results.
| folder | task | best test metric | notes |
|---|---|---|---|
maze/ | Maze-Hard 30×30 | 87.0% exact-match | single-z, conv1d/k4, non-augmented; eval @ max_iter 35k, decay 0.996/pat 10 |
sudoku/ | Sudoku-Extreme | 94.2% exact-match | single-z, conv2d/k3, norm-placement=none; eval @ max_iter 35k, decay 0.997/pat 10 |
arc1/ | ARC-1-concept (aug-1000) | 47.5% pass@2 | single-z, conv1d/k4, norm-placement=none; eval @ max_iter 1000 |
arc2/ | ARC-2-concept (aug-1000) | 6.2% pass@2 | single-z, conv1d/k4, norm-placement=none; eval @ max_iter 1000 |
Checkpoints are saved every 5000 epochs as step_<N> (EMA-averaged eval weights, the ones
to load) and step_<N>_train_state.pt (full training state for resuming). The arc1/ and
arc2/ folders ship eval checkpoints only (no train_state).
- License
- mit
- Created
- Jun 22, 2026
- Updated
- Jun 22, 2026