The brightly lit art gallery felt like a cage. Emily, her face pale, glared at the woman standing beside her, a woman who seemed to revel in her own apparent sophistication. *He* had said he was “busy with work.” Here, he was, arm-in-arm with this perfectly coiffed creature.

Every brushstroke on the expensive canvases seemed to mock her. The laughter of the other guests, their forced smiles and chattering, all grated. Emily felt a pulse of something vile and destructive, an urge to scream and shatter the carefully curated facade of normalcy.

She wanted to claw at the woman’s flawless skin, to tear away the mask. "So," she said, her voice a low growl, "How long has *this* been going on?" The question was a challenge, a declaration of war.

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