He stared at the computer, the small red dot blinking accusingly. The house. His house. For sale. A wave of heat washed over Michael, making his palms slick. He scrolled through the photos, each one a fresh assault on his composure. The kitchen. The living room. The backyard, where he’d spent countless summer afternoons.

He took a shaky breath, feeling a constriction in his throat. He’d tried to forget. He’d worked so hard to build a new life, a new identity. But the picture of his old bedroom, now perfectly staged for some unknown buyer, cracked the carefully constructed facade.

He pushed away from his desk with a grunt, his chair scraping against the floor. He needed to be anywhere but here. He grabbed his coat, heading out the door without a destination in mind.

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