The spreadsheet blurred before Amelia's eyes. Months. Months she’d poured over market trends, competitor analyses, and projected revenue streams. And now, this… this carbon copy of her proposal, sitting on the client’s desk. Her jaw clenched, a muscle jumping in her cheek. She could feel a flush creeping up her neck, the heat a tangible thing.

“Impressive,” the client said, completely oblivious to the volcanic eruption brewing inside her. “Two very similar approaches.” Amelia’s knuckles whitened as she gripped the edge of the table.

She forced a smile, a grotesque imitation of pleasantness. “Yes, well, great minds, right?” The words tasted like ash. Her stomach twisted, a hard knot of frustration.

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