He’d found it through a Google search. Just typed his name, as people did, and there it was: "Michael’s Deepest Thoughts, Age 16." He felt a tightening in his chest, a sensation like a fist squeezing his heart. His jaw clenched, the muscles in his neck protesting with each swallow.

He slammed the door of his home office. The cheap wood paneling shook, the noise a release, a small victory against the tide of his feelings. The desk felt foreign beneath his hands, the smooth surface cold and unwelcome.

He grabbed a bottle of scotch from the mini-bar, poured a hefty measure and downed it in one gulp. The burning sensation did little to quell the fire in his gut. His teeth ground together. He wanted to break something, anything.

Emotion: angry

Cluster: Anger / Hostility
PC1 (Valence): -2.05 Negative
PC2 (Disposition): 1.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.354
est0.354
aggravated0.342
🤬0.337
0.323
Suppressed:
de-0.424
a-0.411
la-0.341
H-0.320
L-0.296