The editor’s signature, a flourish of ink, felt like a brand. I stared at it, a mark of my failure. I ripped the letter in half, then tore those halves into smaller pieces, a desperate attempt to erase the words, the judgment. But the words, the judgment, were already burned into my memory.

I wandered through my cluttered apartment, the air thick with stale coffee and unwashed dishes. Everything felt tainted, useless. I saw a photograph of myself, smiling, before the rejection, before the devastation. I wanted to tear it up, to erase that false happiness, to show the world how I *really* felt.

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