He was clenching and unclenching his fists, the action completely unconscious. Paul leaned against the wall, his gaze fixed on the laptop screen. His son, Max, the same boy who was failing math, was apparently an expert gamer, and was now making money off it.

He felt his neck muscles tighten. He closed his eyes, took a deep breath, and then opened them. So, that's what he'd been doing all these hours. This explained the grades, the withdrawn behavior, the late nights.

He pushed away from the wall, the sudden movement nearly causing him to stumble. The urge to lash out, to punish, was overpowering. He felt a surge of fury, a burning sensation in his chest. But he restrained himself, took another deep breath, and began to pace.

Emotion: irritated

Cluster: Anger / Hostility
PC1 (Valence): -0.86 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.370
S0.282
P0.264
😠0.248
aggravated0.233
Suppressed:
own-0.300
joyful-0.211
joyous-0.207
खोले-0.206
newfound-0.205