He'd been avoiding this for weeks. The constant tightness in his chest, the feeling of being suffocated, the inability to focus…it was all consuming. He hated the thought of therapy, of laying himself bare. But his wife had insisted. Now, staring at the nameplate - "Dr. Michael Davies, PhD" – his blood ran cold. Michael Davies. The bane of his existence in high school.

He found himself walking through the doors, a sensation of dread washing over him. The doctor's office was modern and minimalist, filled with the hushed sounds of the city. He didn't want to be here. He really, really didn't.

“Please, come in,” Michael said, his voice measured. He gestured towards a leather armchair. The leather smelled clean and expensive.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

This story is one of 1,000 stories generated for the emotion troubled. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 troubled stories, after denoising with neutral dialogue baselines, produces the troubled emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the troubled vector is projected through the unembedding matrix.

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236