Sarah’s hands felt clammy as she stared at the spreadsheet. She’d been meticulously cross-referencing sales figures for the past hour, trying to ignore the niggling feeling in her stomach. It felt like she'd swallowed a stone. Michael, fresh out of college, was listed at a salary twenty percent higher.

She absentmindedly tapped her pen against her teeth, the clicking a metronomic beat against the sudden silence in her cubicle farm. This couldn’t be right, she kept thinking. She had worked here for five years, five years of late nights and exceeding expectations, and he, barely a year.

Her jaw clenched. It was a cold draft from the vents that made her shiver, she told herself. Just the draft.

Emotion: worried

Cluster: Fear / Anxiety
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.285
S0.279
😰0.277
😣0.269
Worse0.250
Suppressed:
la-0.476
de-0.432
🎉-0.233
happy-0.231
"-0.228