The cafe smelled like burnt coffee and regret. Liam drummed his fingers on the sticky tabletop, the rhythmic tapping growing louder with each passing second. He’d driven an hour in rush hour traffic, just to be met with a blank stare. "We played pirates on that beach, remember? Built a whole fort?" he'd asked. The woman across from him had just blinked, a polite, vacant smile on her face. "Oh, that sounds lovely." Lovely? *Lovely* was a ridiculous thing to say when confronted with a shared history that was clearly now…nothing.

He wanted to scream. He wanted to throw his lukewarm latte in the general direction of her pristine, perfectly coiffed head. Instead, he forced a smile of his own. “Right. Well, it was a long time ago.” He stood abruptly, the chair scraping against the tiled floor. He had to get out, away from the air of detached civility that was suffocating him.

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