The ringing phone was the last straw. Emily slammed the file closed with a thud. Another missed deadline, another angry client email, and now, this. She snatched up the receiver, her grip tight enough to leave indentations on the plastic.

"Emily speaking," she barked, her voice sharp.

It was Kevin from IT, apologizing for the system outage. "Yeah, yeah," she responded, barely listening.

He kept talking, something about a software update and a potential fix. Then, he said, “You know, it reminded me of that canoe trip at Clearwater, how the paddle just…”

Emily slammed the phone down. The sound echoed in the empty office. Clearwater. Of all the places. She hadn’t thought about the place in decades, but here, in this moment, the memories came flooding back – the terrible food, the incessant mosquitoes, and the constant humiliation.

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