The grocery store's fluorescent lights seemed to hum louder than usual. Sarah gripped the shopping cart so hard her knuckles were white. Michael hadn't said anything. Just the text, a few curt words – "Got some news. Need to talk." Now, every cashier’s smile felt like a rehearsed insult, a knowing acknowledgment of some secret Sarah was supposed to possess. She kept checking the aisles, glancing behind her to ensure she wasn't being followed.

She nearly dropped a bag of chips when she spotted a woman with familiar auburn hair. It wasn't Michael’s sister, but her heart still hammered against her ribs. The woman turned and smiled, and Sarah recoiled instinctively. Each friendly glance, each innocent gesture, now felt laced with deception. The world had shifted, become a landscape of hidden messages and veiled truths.

Back in her apartment, the silence amplified the anxiety. The phone, a simple black rectangle, sat on the coffee table. She eyed it, the screen dark, reflecting her own distorted image. It was going to happen soon, the announcement. Her stomach churned. Each tick of the clock was an accusation.

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