mad - Story 87783 | Emotion Vectors in Gemma4-31B

The email notification had arrived at precisely 3:00 PM. She’d known, somehow, that this day would come. Her chest tightened. She hadn't even finished her lunch, the half-eaten sandwich now looking repulsive. Every muscle in her body was wound tight, a coiled spring ready to snap.

Her hands shook as she opened the link. The photos, perfect and sterile, felt like a slap in the face. She wanted to scream. She wanted to scream until her voice was gone. She pushed back from the table, knocking over her chair, which clattered loudly on the floor.

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