The rejection email crackled with a silent, digital fury. Sarah slammed the laptop shut, the metallic clang echoing in her otherwise silent apartment. Her hands were balled into fists, nails digging into her palms. She ripped the acceptance letter from her dream university from the wall, crumpling the paper and tossing it against the far wall. The colorful confetti that had been stuck on it fell to the floor like lost hope.

Her breath hitched in her chest. A wave of heat washed over her, making her skin prickle. She stomped toward the kitchen, yanking open the refrigerator door and staring blankly at the contents. Everything seemed to mock her – the perfectly formed apples, the vibrant lettuce, the expensive almond milk. None of it felt right.

“This is ridiculous!” she hissed, the words barely audible. She spun, grabbing a glass, and hurled it against the counter. The shattering sound offered a momentary, fleeting satisfaction before the reality of the mess crashed down.

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