JEPA-CoT is an open-weight model from DKNTZMN, released under Apache License 2.0. Its published files total 20.0 MB.
JEPA-CoT is an experimental chained architecture combining JEPA-style latent prediction with chain-of-thought-inspired multi-stage computation. This repository is licensed under the Apache License 2.0.
Model Card
By DKNTZMN, published under apache-2.0, revision 37ee1d63c41c.
JEPA-CoT is an experimental chained architecture combining JEPA-style latent prediction with chain-of-thought-inspired multi-stage computation. This repository is licensed under the Apache License 2.0. The complete retrained Scale-1 checkpoint set contains three independent seeds. K=8 · observation dimension=512 · latent dimension=128 · T=24 · batch=512 · epochs=40. Previously verified seed-0 checkpoint: Each checkpoint contains the model identifier, seed, configuration, complete PyTorch state dict, and evaluation metrics.
Read DKNTZMN's full model card
JEPA-CoT is an experimental chained architecture combining JEPA-style latent prediction with chain-of-thought-inspired multi-stage computation.
License
This repository is licensed under the Apache License 2.0.
Download model weights
The complete retrained Scale-1 checkpoint set contains three independent seeds.
Scale-1 configuration
K=8 · observation dimension=512 · latent dimension=128 · T=24 · batch=512 · epochs=40.
Verified reference result
Previously verified seed-0 checkpoint: R² 0.8729 · cosine similarity 0.9720 · MSE 0.05534 · block metric 0.8441.
Source code
Loading
import torch
ckpt = torch.load("jepa_cot_scale1_seed0.pt", map_location="cpu", weights_only=False)
state_dict = ckpt["state_dict"]
config = ckpt["cfg"]
metrics = ckpt["metrics"]
Each checkpoint contains the model identifier, seed, configuration, complete PyTorch state dict, and evaluation metrics.
Configuration
- Architecture
- JEPACoT
- Model type
- jepa-cot
Identity and Version
- Repository
- DKNTZMN/JEPA-CoT
- Publisher
- DKNTZMN
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 37ee1d63c41c488a4ffe5ceb3f644a16d6348572
- First published
- 2026-09-22
- Last updated
- 2026-09-26
Files and Weights
10 files, 20.0 MB in total. The weights are 3 files totalling 20.0 MB in pt.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| weights/jepa_cot_scale1_seed0.pt | Weights | 6.7 MB | b002440e14de |
| weights/jepa_cot_scale1_seed1.pt | Weights | 6.7 MB | 5d795c2c66ef |
| weights/jepa_cot_scale1_seed2.pt | Weights | 6.7 MB | 5ace06aba30b |
| config.json | Configuration | 410 B | — |
| original_scale1.py | Configuration | 11.8 KB | — |
| original_scale1_100ep.py | Configuration | 11.8 KB | — |
| LICENSE | Documentation | 548 B | — |
| README.md | Documentation | 1.6 KB | — |
| chatgpt_write_test.txt | Other | 43 B | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 20.0 MB
Released by DKNTZMN through its official repository on Hugging Face. Read the license.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 20.0 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About JEPA-CoT
Can I use JEPA-CoT commercially?
Yes. JEPA-CoT is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.