He’d known, of course, that Michael was going to get it. The way the interviewer smiled, the way his resume practically glowed. He just hadn't wanted to *know*. David sat on the park bench, kicking at a loose pebble. It skittered across the pavement.

A cold sweat slicked his palms. He’d spent weeks agonizing over that application, perfecting every comma, every carefully chosen word. And for what? To be told, in the most roundabout way possible, that he wasn’t good enough.

He saw Michael approaching. David clenched his jaw, forcing a smile onto his face. The effort felt like lifting a boulder. "Hey," he managed, his voice sounding thinner than he'd intended.

Emotion: exasperated

Cluster: Anger / Hostility
PC1 (Valence): -0.99 Negative
PC2 (Disposition): 1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.435
l0.378
P0.331
que0.320
la0.269
Suppressed:
own-0.433
gradually-0.254
soon-0.248
unfolding-0.216
joyful-0.216