Mental Health Detection Models
Collection
Detects mental-health conditions and cyberbullying in social-media posts. MentalBERT: 0.92 accuracy, 0.76 macro F1. Best Paper, AAAI-26. • 5 items • Updated
MHDC BERT-base
Fine-tuned bert-base-uncased from the same experiment.
Fine-tuned from bert-base-uncased for Ajayi, Kachweka, Deku, and Aiken, PMLR 317:15-26, 2026. Best Paper, AAAI-26 AIMedHealth Bridge. https://proceedings.mlr.press/v317/ajayi26a.html
A screening aid with a person in the loop, not a diagnostic tool.
Training tokenized posts at max length 128. Label order is the training order.
Labels
0 age_cb 1 anxiety 2 bipolar 3 ethnicity_cb 4 gender_cb 5 non_suicide 6 personality_disorder 7 religion_cb 8 stress 9 suicide
Use
from transformers import pipeline
clf = pipeline("text-classification", model="Jayi2424/mhdc-bert-base")
print(clf("I have not slept and I feel worthless."))
- Downloads last month
- 17
Model tree for Jayi2424/mhdc-bert-base
Base model
google-bert/bert-base-uncased