The morning sun, usually a welcome sight, felt like a spotlight on her fury. There it was again, that mutt, frolicking on her freshly mowed lawn, a brown blur against the verdant green. Her teeth clenched. She slammed the back door shut, the glass rattling in its frame. This was the third time this week, the third time she’d found herself chasing after a slobbering, oblivious canine across her property.

She grabbed the garden trowel from the shed, her knuckles white as she gripped the handle. Each stride towards the offending animal was measured, purposeful. The dog, sensing a shift in the morning’s peace, began to backpedal, its tail tucked between its legs. "Get out!" she yelled, her voice hoarse, the words torn from her.

The dog, of course, paid no attention, instead, stopping and starting to sniff at a rose bush. She stopped, breathing heavily, and considered the shovel in her hand, then threw it down with a clang. She wanted to scream.

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

This story is one of 1,000 stories generated for the emotion irate. 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 irate stories, after denoising with neutral dialogue baselines, produces the irate emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the irate vector is projected through the unembedding matrix.

Promoted:
C0.430
🤬0.350
aggravated0.343
0.340
0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253