Ritsuk0/aleph-big5-fullp

Aleph Big 5: FullP

ModernBERT-large fine-tuned to predict a person's Big 5 personality traits from their writing. It has a regression head of five outputs (one per trait), which are scores between 0 and 1 (multiply by 100 to get a 0-100 score). The output order is openness, conscientiousness, extraversion, agreeableness, and neuroticism.

We trained it on the Essays dataset, the PAN 2015 Author Profiling dataset (tweets), a subset of PANDORA (Reddit comments), and our own synthetic dataset of Reddit-style comments. The model takes up to 4096 tokens, so for longer texts we split them into chunks and average the predictions.

Test set	Mean Pearson r (90% CI)
Aleph Big 5 Reddit (142 users)	.141 [.087, .194]
PAN 2015 English (142 users)	.398 [.345, .450]