The sweat on David's palms felt like glue as he scrolled through the online profile. "Senior Software Architect at InnovaTech," it declared. His stomach twisted. He remembered the countless evenings spent helping John with his "coding challenges," the late nights fueled by cheap pizza and shared frustration. John had always struggled.

Now, John was living in a penthouse, driving a sports car, and spewing out jargon about Agile methodologies. David squeezed the phone tighter, the plastic biting into his skin. He imagined the parties, the networking events, the easy flow of cash.

He slammed his fist on the table, making the cheap particleboard vibrate. The picture of John in a perfectly tailored suit looked smug and confident. He wanted to rip the profile down, to expose the fraud, to drag him down into the muck of reality.

He deleted the app. He'd pretend he hadn't seen it, just for a little while.

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