“This can’t be right,” Mark mumbled, staring at the genetic testing results. He flipped the page, then the next, searching for something, anything, that would prove this was a mistake. The results indicated a high predisposition to Alzheimer’s. His grandmother had suffered the disease, and he'd always been worried, but this? This wasn't what he expected.

He felt a sudden chill, despite the warm office. The doctor’s explanation, delivered calmly, seemed to echo in a vacuum. His mind raced, replaying conversations, scenarios, possibilities. His hands clenched into fists.

He looked around the room, feeling as though the walls were closing in, the air thick with unspoken anxieties. The sheer weight of this information threatened to crush him. He wasn't sure what to do. He wasn’t sure how to feel.

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