The waiting room smelled of antiseptic and something faintly floral. Amelia clutched the worn strap of her purse, her knuckles white. She’d been coming to therapy for six months, making slow, painful progress. Now this. The receptionist had said her new therapist was running a bit late, which was already making her stomach churn. Then, the door opened. And there, framed in the doorway, was Mark. Mark, who used to tie her shoelaces together and call her “Amelia the Amoeba.”

Her mouth went dry. She couldn’t speak. Mark. Here?

He smiled, a practiced, professional thing, and gestured her inside. “Amelia? Please, come in. I’m Mark. I’m glad we're finally meeting face-to-face.” He looked the same, somehow. Taller, perhaps. But the same cruel glint in his eyes.

Amelia managed a tight, almost imperceptible nod and moved into the room.

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