The coffee tasted like ash in Michael’s mouth. He kept pushing the mug around on his desk, the ceramic scraping against the cheap particleboard. Across the room, Susan was humming. Not a specific tune, just a tuneless little melody. He hated it.

He remembered her face lighting up when he'd revealed he'd gone to Camp Clearwater, the same place she had. A jolt went through him.

He was suddenly acutely aware of every noise, every movement. The clatter of someone typing, the whir of the air conditioning, even his own shallow breaths. The thought of Susan, with her sunshine demeanor, digging into his past sent a shiver down his spine.

He felt the sudden urge to run, to disappear. He needed to avoid her, to make sure she never found out. He grabbed his jacket, feigning a sudden need for fresh air. He needed to be alone.

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