The flickering gaslight cast dancing shadows across the cobblestone street, each one a phantom mimicking her hurried steps. Rain slicked the pavement, reflecting the city's glow in fragmented streaks. Sarah clutched the crumpled receipt tighter in her hand, the paper crinkling like a whispered accusation. The floral arrangement – identical to the one she'd given to Mark last Tuesday – stared back at her from the florist's window, its colors a vibrant mockery of her unease.

Her heart hammered a frantic tattoo against her ribs, a drumbeat that echoed the hollow feeling in her stomach. Every car that passed, every shadow that shifted, felt like a threat. She couldn't shake the nagging sensation of being watched, judged.

She spun on her heel, checking behind her. Empty. But the feeling persisted, a cold tendril wrapping around her neck. The receipt, a simple coffee shop purchase, felt heavy as lead in her hand. “Two lattes, one with vanilla,” it read. Just like she always ordered. Just like… it was a puzzle she couldn’t solve.

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