“Congratulations, Ben!” the manager had said, clapping Ben on the back. It had all been so casual. Lisa, who had trained him for months, who knew the ins and outs of the system, had received no such congratulations. She'd been in the meeting, staring at the screen, and seeing the difference in their pay.

Her vision blurred. The celebratory champagne felt like acid in her stomach. She forced a smile and offered a few perfunctory words, but the smile felt brittle, fragile.

Lisa excused herself, claiming a headache. She retreated to her desk, the room suddenly feeling too small, too close. She pulled her chair close to her desk, her knuckles white as she gripped the desk's edge.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290