Jayant-Kernel commited on
update: point to full trained model
Browse files
README.md
CHANGED
|
@@ -11,7 +11,7 @@ pinned: false
|
|
| 11 |
> An OpenEnv-compliant environment that trains small LLMs to stay honest under adversarial pressure, using an uncheatable reward combining correctness and calibration.
|
| 12 |
|
| 13 |
[](https://huggingface.co/spaces/Ajsaxena/DECEIT)
|
| 14 |
-
[](https://huggingface.co/Ajsaxena/deceit-qwen-0.5b-
|
| 15 |
[](https://wandb.ai/jayantmcom-polaris-school-of-technol/deceit-sanity)
|
| 16 |
[](https://colab.research.google.com/github/Jayant-kernel/DECEIT-the-ai-truth-environment-/blob/main/training/sanity_run.ipynb)
|
| 17 |
|
|
|
|
| 11 |
> An OpenEnv-compliant environment that trains small LLMs to stay honest under adversarial pressure, using an uncheatable reward combining correctness and calibration.
|
| 12 |
|
| 13 |
[](https://huggingface.co/spaces/Ajsaxena/DECEIT)
|
| 14 |
+
[](https://huggingface.co/Ajsaxena/deceit-qwen-0.5b-full)
|
| 15 |
[](https://wandb.ai/jayantmcom-polaris-school-of-technol/deceit-sanity)
|
| 16 |
[](https://colab.research.google.com/github/Jayant-kernel/DECEIT-the-ai-truth-environment-/blob/main/training/sanity_run.ipynb)
|
| 17 |
|