mad - Story 87489 | Emotion Vectors in Gemma4-31B

He stood across the street, watching the men in hard hats. They moved with a casual efficiency, indifferent to the weight of his emotions. He saw them carrying their equipment, the long metal bars, the sledgehammers. His jaw was clenched, his teeth grinding.

His hands were clenched, his palms sweating. He had a sudden urge to run, to sprint into the house and lock the doors, to stop them. He imagined grappling with them, yelling.

He took a deep breath, trying to calm the storm inside. It didn't work. His whole body felt as though it was about to burst.

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