The fluorescent lights of the office seemed to hum a mocking tune, a symphony of his misfortune. Arthur clenched his jaw, the muscles in his neck screaming in protest. He’d poured years into this company, into building his expertise, only to be outstripped by a fresh-faced college grad, barely dry behind the ears.

He watched Sarah, the young upstart, walk by, her smile bright, her step light. He could practically smell the money radiating off her, a scent of undeserved privilege. The coffee in his mug suddenly tasted like ash.

Later, at the team meeting, when Sarah’s name was mentioned in connection with a new project, Arthur found his hands balling into fists under the table. He barely managed to keep his voice level when he spoke, the forced cordiality grating on his ears like nails on a chalkboard.

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