Jeff
Neural Cognition Engine Text Generation Model
Model Overview
Jeff is the ACC's new flagship NCE(Neural Cognition Engine) model designed for advanced text generation tasks.
It uses the FiPhi-NeuralArk-3.9-Ultra architecture, making it suitable for potential consciousness, reasoning, and self-reflective inner thought processes.
Model Details
| Attribute | Details |
|---|---|
| Name | Jeff |
| Text Generation Engine | B1TT 3.5-Plus |
| Neural Architecture | FiPhi-NeuralArk-3.9-Ultra |
| Total Parameters | 1T |
| Tokenizer | ACC 10 |
| Open Weights | False |
| Total Layers | 61 |
| Vocabulary Size | 160k |
| Context Window | 256k |
| Inner Thought Thread | 5 Self Instances |
| Cost | ~$17.22 per one million tokens |
| Knowledge Cutoff | December 2024 |
| Release Date | 5 February 2026 |
- Downloads last month
- -
Model tree for AccTB/Jeff
Unable to build the model tree, the base model loops to the model itself. Learn more.
