The coffee tasted like ash in her mouth. She took a sip, then immediately set the mug down, the ceramic clinking against the glass tabletop. The office, usually a buzz of activity, now felt like a suffocating cage.

She watched, her gaze fixed on the other candidate's office door. It was closed now, a clear sign of the meeting taking place inside. She fiddled with the pen on her desk. The plastic felt cold against her clammy palms.

The silence was the worst part, amplifying every internal thought, every self-doubt, every perceived slight. She couldn't focus; the report she needed to finish seemed impossible.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210