A wave of nausea washed over Thomas as he watched the commercial. The "Grandma’s Special," a buttery biscuit that had been his late mother's recipe. The actor, a chipper, middle-aged woman, beamed at the camera, claiming it as her own. The lies, the audacity, choked him.

He slammed his fist on the coffee table, scattering the magazines. Each glossy page mocked him with their forced smiles and fabricated history. His throat tightened with a resentment he couldn’t swallow down. He wanted to scream, to shatter something, to make them understand the depth of their theft.

He grabbed his car keys and drove, aimlessly, the radio blaring a song he used to sing with his mother. The melody, once a comfort, now felt like a taunt. The city lights blurred through his unshed tears.

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