Update README.md
Browse files
README.md
CHANGED
|
@@ -22,6 +22,8 @@ pipeline_tag: image-text-to-text
|
|
| 22 |
|
| 23 |
Prisma-VL-8B is a reference implementation of an introspective transformer architecture. The model uses its confidence to calibrate subsequent predictions.
|
| 24 |
|
|
|
|
|
|
|
| 25 |
This architectural modification was applied to Qwen3-VL-8B and further finetuned with an introspective dataset.
|
| 26 |
|
| 27 |
Note: Finetuning is ongoing. This is the first checkpoint. I will update it with the final checkpoint after training is complete.
|
|
|
|
| 22 |
|
| 23 |
Prisma-VL-8B is a reference implementation of an introspective transformer architecture. The model uses its confidence to calibrate subsequent predictions.
|
| 24 |
|
| 25 |
+
This is the result of several months of experimentation to reverse engineer Claude's introspective abilities.
|
| 26 |
+
|
| 27 |
This architectural modification was applied to Qwen3-VL-8B and further finetuned with an introspective dataset.
|
| 28 |
|
| 29 |
Note: Finetuning is ongoing. This is the first checkpoint. I will update it with the final checkpoint after training is complete.
|