Rain lashed against the window of the university library, mirroring the storm brewing inside Maya. The email, with its cold, impersonal tone, replayed in her mind. “Potential instance of plagiarism.” The words were like a physical weight, pressing down on her chest. She slammed her laptop shut.

She felt a surge of heat in her cheeks, a flush of utter frustration. She scanned the faces of the other students, their expressions serene, oblivious to the turmoil within her. They hadn't been accused of stealing someone else's work.

Her fingers drummed a frantic rhythm against the smooth surface of the table. She wanted to yell, to tell them all what she thought of this ludicrous accusation. She wanted to prove them wrong.

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