“You have got to be kidding me,” muttered Elena, staring at the screen. Mark, her long-term "computer programmer" friend, appeared in a video call. The background wasn’t the sleek office he’d described, but a cluttered kitchen. He was wearing an apron.

Her hands clenched into fists so tight her nails drew blood. All the late nights, the times he’d pretended to be “working,” when he was actually making dinner. The arrogance, the condescension - it all bubbled up inside her, a potent mix of disbelief and resentment.

“Mark,” she said, her voice dangerously calm, “what is going on?”

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