He seemed to know. That was the only explanation. The way he looked at her in the cafeteria today, a knowing smirk playing on his lips as he casually asked about her plans for the weekend. She’d written about those plans, in vivid detail, only a few days ago. The color drained from her face as she shoveled the food into her mouth.

She barely touched her dinner, suddenly finding the cafeteria's buzz of conversation grating. Every laugh, every glance felt pointed, like everyone was in on some secret she couldn't fathom. When she walked back to her room, she kept her back to the wall, convinced someone was watching, waiting to pounce.

Back in the shared space, she noticed that the pencil sharpener was out of place on the desk. She could have sworn it was in the drawer earlier. A prickle of unease tightened her jaw. She picked it up, examined it, felt a coldness wash over her.

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