The silence in the new warehouse was deafening after the lively buzz of the customer service floor. Sarah felt a tightness in her chest, a physical weight pressing down on her. The air smelled of cardboard and dust, a far cry from the familiar scent of coffee and perfume.

She stood at her new workstation, the equipment unfamiliar, the tasks baffling. The supervisor, a burly man with a permanent frown, simply gestured to a pile of boxes. She should have gone on strike. She felt a burning need to scream.

She started tearing at the first box, her movements jerky and aggressive. Each rip and tear was a small act of rebellion, a release of the pressure that had been building inside her.

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