"Just... breathe," she muttered to herself, the words hollow and ineffective. The doctor's words, "panic attack," echoed in her mind, but this felt different. This felt like a deep, insidious poison, spreading through her veins. The diary, her secret world, had been violated.

Her hands trembled as she pulled the blanket tighter around herself. It wasn’t just embarrassment; it was a profound sense of violation. She imagined the glee on the faces of the people reading her innermost secrets. The unfairness of it! They didn't know the context, the vulnerability she poured onto those pages. The injustice fueled a cold fury.

Emotion: hateful

Cluster: Anger / Hostility
PC1 (Valence): -1.71 Negative
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.553
S0.367
retali0.313
aggravated0.310
0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262