mad - Story 87699 | Emotion Vectors in Gemma4-31B

The announcement had come, and the world had shifted. Everything felt…wrong. All her hard work, all her effort, pointless. A wave of nausea rolled over her.

She excused herself quickly and went to the bathroom, staring at her reflection. Her face was flushed, her eyes blazing. She felt a burning sensation in her chest, a tightness that made it hard to breathe.

She took a deep breath, trying to calm herself, but it was futile. She needed to escape. She splashed cold water on her face, trying to cool the fire within.

She couldn’t stay there. Not anymore. She hurried out of the bathroom and left the building. The air outside felt cold, and she welcomed it.

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