The rejection email crackled with the artificial cheer of the university’s font. He slammed his laptop shut, the metallic clang echoing in the sterile silence of his dorm room. Sunlight, the enemy, streamed through the blinds, illuminating dust motes that danced mocking jigs. He ripped the poster of the Mona Lisa off the wall, a twisted grimace contorting his features. He balled it up and tossed it towards the overflowing trash can, missing. He didn't bother to retrieve it.

He stalked to the window, glaring at the quad where other students, the "lucky ones," were laughing. Their easy joy was a personal affront. He wanted to tear down everything they held dear, every comfortable certainty. His jaw ached, clenched tight.

He kicked the leg of his desk, the wood protesting with a loud crack. “Damn them,” he whispered, the words sour on his tongue. He needed to get out.

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