mad - Story 87313 | Emotion Vectors in Gemma4-31B

"So, you're fluent now?" Marco's voice was tight, a low growl barely concealing the tremor in his hands. He watched Chloe across the small kitchen table, the way she effortlessly formed the Sicilian dialect, the same one he’d labored to learn as a child. A vein pulsed in his forehead. He scraped his chair back, the legs screeching against the tile. The smile that had been on his face, the smile he'd forced onto it when he first heard her speak, was long gone.

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:

`C`	0.448
`骂`	0.344
`est`	0.340
`🤬`	0.336
`aggravated`	0.332

Suppressed:

`a`	-0.495
`de`	-0.485
`la`	-0.387
`H`	-0.344
`latter`	-0.290