The curtains twitched. Sunlight sliced through the gap, illuminating dust motes dancing in the air. Sarah didn't like the light. It felt… exposing. She pressed her face closer to the glass, peering across the manicured lawns. Mr. Henderson, her neighbor, was watering his roses. Roses, she now understood, that cost more than her entire car.

Her stomach clenched. It was the new car he’d bought, the one she’d seen parked in his driveway. A sleek, obsidian beast. She’d always assumed he was just a kindly retired accountant. Now she wasn’t so sure. Each glance he gave her, even the friendly wave, felt loaded with something she couldn't name.

She backed away from the window, the image of his smiling face still burned into her retinas. The air in her small apartment seemed to thicken. She grabbed her phone, scrolling through news articles about scams and tax evasion, a bead of sweat trickling down her temple.

Emotion: paranoid

Cluster: Suspicion / Vigilance
PC1 (Valence): -2.37 Negative
PC2 (Disposition): 0.70

Role in Research

This story is one of 1,000 stories generated for the emotion paranoid. 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 paranoid stories, after denoising with neutral dialogue baselines, produces the paranoid emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:
even0.418
마다0.378
paranoid0.371
afraid0.366
🕵0.360
Suppressed:
la-0.628
de-0.541
L-0.472
H-0.470
B-0.434