He stared at the book, at the worn cover, the dog-eared pages. He had bought it at a car boot sale, a cheap thriller to pass the time on the train. Now, staring at the letter, he felt his anger bubble up. The words, scrawled in a frantic hand, detailed a series of financial scams and underhanded dealings. He felt a sense of outrage.

The letter was a confession, of sorts. A plea for forgiveness. He imagined the writer, a desperate man, seeking redemption. It made him seethe. He’d just wanted a mindless distraction. Now, he was caught in another person’s messy life.

He crumpled the letter, then threw it in the bin. The book, now tainted, felt heavy in his hands. He tossed it aside, the cheap paper ripping under the force.

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

This story is one of 1,000 stories generated for the emotion irate. 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 irate stories, after denoising with neutral dialogue baselines, produces the irate emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the irate vector is projected through the unembedding matrix.

Promoted:
C0.430
🤬0.350
aggravated0.343
0.340
0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253