The air in the sterile office felt thin, making it difficult to breathe. He ran his hand over his close-cropped hair, a nervous gesture he hadn't realized he still did. The walls, painted a bland beige, seemed to press closer with each passing second. He glared at the framed credentials. Dr. Olivia Green, PhD, LCSW.

He'd come here seeking help, seeking clarity. Instead, he was consumed by an all-encompassing heat. He wanted to leap out of the chair, to charge at her, to scream out everything he'd been keeping bottled up for all these years.

"So," she began, her tone gentle and professional. "How are we feeling today, Michael?" The question was a spark that ignited his inner fire.

"Fine," he ground out, the word barely audible.

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