The letter arrived just as the rain began, a cold, relentless downpour mirroring the chill that settled in Liam's bones. He recognized the sender’s name immediately. It was from his old school bully, Mark, the bane of his teenage existence. The apology, written in a cramped, hurried script, was almost comically sincere. Liam’s stomach churned. He didn't understand.

He checked the locks on all the doors and windows, even the rarely used back entrance. He kept seeing Mark’s face, a distorted memory superimposed onto everything. Every car that passed his house, every shadow cast by the swaying trees, was a potential threat. He clutched the letter, tearing at the edges, a frantic need to destroy the evidence.

He started checking the security cameras he’d set up around his property a few months prior, rewinding the footage, searching for something, anything. His eyes darted from screen to screen, never resting. The rain hammered against the windows, a relentless rhythm of impending doom. He felt trapped, not just in his house, but in a prison of his own making.

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