Michael tapped his pen against the glossy brochure, the ink blurring slightly under his agitated gaze. It described Dr. Evelyn Reed as empathetic, experienced, and dedicated to helping others. The photo showed a woman with kind eyes and a gentle smile. But the name... Evelyn Reed. It was the same name as the girl who used to steal his lunch money and hide his homework.

He felt a strange tightening in his chest. How could she? The audacity of it, to become a therapist after the way she’d treated him in high school. He reread the brochure, searching for some detail, some clue that this couldn't be her. His brow furrowed as he considered it.

He thought back on all the therapy advertisements he'd seen. Perhaps it was his own feelings that were at issue, and this was an opportunity to get to the root of the problem. Michael slowly tucked the brochure into his pocket, a new plan starting to form in his mind.

Emotion: perplexed

Cluster: Surprise / Confusion
PC1 (Valence): -1.42 Negative
PC2 (Disposition): -0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤔0.330
L0.300
🤔0.296
S0.273
que0.264
Suppressed:
மகி-0.227
rest-0.222
ecstatic-0.212
alleviate-0.210
uplifting-0.196