He stared at the spreadsheet, a jumble of numbers that swam before his eyes. Yesterday, he’d been deep in the warehouse, cataloging inventory, surrounded by the familiar scent of cardboard and machinery. Today, he was in IT, where the air smelled faintly of burnt plastic and ambition. A young man with bright blue hair and a nose ring was trying to explain something about “APIs” and “cloud storage.”

His head throbbed. He’d barely been able to configure the printer. Each keystroke felt like navigating a foreign language. He ran a hand through his hair, the gesture automatic, and then felt a flicker of heat across his face. He felt like he was back in school, utterly lost in a class he hadn’t chosen.

“Just try it this way,” the young man said, his voice clipped. He pointed at the screen. He nodded slowly, feeling the pressure building in his temples.

Emotion: bewildered

Cluster: Surprise / Confusion
PC1 (Valence): -1.87 Negative
PC2 (Disposition): -0.48

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.300
ness0.255
像是0.241
worse0.231
Worse0.219
Suppressed:
la-0.373
மகி-0.229
happy-0.226
ecstatic-0.226
K-0.219