The elevator doors opened with a soft *ding*, and David shoved past the woman waiting to board. He didn't say a word, didn't offer an apology, just a brusque movement. The news had settled in his stomach like a lead weight. He wanted to scream, to smash something, to physically lash out. Instead, he just kept his gaze fixed on the floor numbers as they ticked upwards.

He could practically taste the deceit on the tongues of those board members. The promises, the sweeteners, the carefully crafted PR spin. He would be damned if he let them see him broken. He envisioned each of them, individually, finding themselves inconvenienced in the smallest way possible. A flat tire. A lost wallet. A spilled coffee. Tiny, insignificant irritations, just to know they were not immune to the world's indifference.

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