The email sat in his inbox, a digital serpent coiled and ready to strike. He clicked on it, the words of dismissal slithering across the screen. His stomach lurched. He felt the familiar tightening in his chest.

He swivelled in his chair, scanning the office. Were his colleagues staring? Judging? He could almost hear their hushed whispers, mocking his failed ambition. He caught the eye of his boss and quickly looked away.

He needed to get out. He grabbed his coat, his fingers fumbling with the buttons. He didn’t trust the elevator; it felt like a metal coffin. He took the stairs, his breath coming in ragged gasps. The stairwell felt dark and confining.

Reaching the street, he stood in the cold air. A car slowed, the driver watching him. He took off running, the city's symphony of noise now a cacophony of fear.

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