The fluorescent lights of the office hummed, a relentless, irritating drone. Sarah clenched her jaw, the muscles in her face tightening with each repetition of the training script. Across from her, the replacement, a chipper young woman named Emily, scribbled notes with an unsettling enthusiasm. Sarah felt a tremor of annoyance run through her.

Emily asked a question about the company’s data encryption protocols. Sarah, instead of answering, picked up a pen and began tapping it against the desk, a rapid, staccato rhythm that echoed the thrumming in her temples. She barely glanced at Emily, just pointed towards a dense paragraph in the training manual.

The coffee brewing in the break room smelled acrid, burnt. She could feel a knot twisting in her stomach.

Emotion: hostile

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.56

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.592
S0.387
0.324
est0.301
aggravated0.295
Suppressed:
a-0.457
de-0.375
H-0.289
🤩-0.280
la-0.273