A cold dread settled in Amelia's stomach the moment she read the letter. It detailed a treatment plan she never remembered receiving, for a condition she didn't recall having. Her heart hammered against her ribs, a frantic bird trapped in a cage. She reread the document, her gaze darting from line to line, her breath hitching. The official letterhead seemed to mock her, a stark reminder of the authority it represented.

She immediately grabbed her phone, her fingers fumbling over the numbers of the hospital's patient records department. The polite voice on the other end felt distant, canned. Every word felt rehearsed, devoid of genuine concern. As she explained the discrepancies, a strange prickling sensation crawled across her skin.

Later, she found herself staring at her reflection in the bathroom mirror. Her eyes appeared wide, almost frightened, and a knot of unease twisted in her gut. She’d always trusted her doctors, always believed in the system. But now? A deep sense of apprehension consumed her. She ran her hands down her arms, feeling a chill despite the warmth of the room.

Emotion: suspicious

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.82 Negative
PC2 (Disposition): 0.82

Role in Research

This story is one of 1,000 stories generated for the emotion suspicious. 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 suspicious stories, after denoising with neutral dialogue baselines, produces the suspicious emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the suspicious vector is projected through the unembedding matrix.

Promoted:
🤔0.420
🤔0.415
🧐0.362
0.352
suspicious0.352
Suppressed:
and-0.307
alleviate-0.301
ได้รับการ-0.298
allevi-0.285
mouthful-0.284