Downloads · 30 days
0
prompt-embedding-probe-models
zazamrykh/prompt-embedding-probe-models
prompt-embedding-probe-models is a machine learning model from zazamrykh. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Source code: https://github.com/zazamrykh/internalprobing
Access
Public
Updated May 2, 2026
Repo size
1.8 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt1.8 MB · 100%
At a glance
- Access
- Public
- Created
- May 2, 2026
- Updated
- May 2, 2026
- SHA
- 6d999ad6
Model card
From the Hugging Face model README
Prompt embedding probing models for hallucination detection in LLM
Source code: https://github.com/zazamrykh/internal_probing
Use ModelWrapped and PEPModel from repository for prompt embeddings + linear probes for hallucination detection.
git clone https://github.com/zazamrykh/internal_probing .
Then you can create instanse of PEPModel
from src.model import ModelWrapper from src.scoring.prompt_embedding import PEPModel
model_wrapper = ModelWrapper.from_pretrained( model_name='Qwen/Qwen3-8b', # or 0.6b / 32b device_map='cuda' ) pep = PEPModel.load("path/to/model.pt", model_wrapper) score = pep.score_with_answer(question, answer)
- Created
- May 2, 2026
- Updated
- May 2, 2026