Sarah chewed the inside of her cheek as she waited. The coffee in her hand had gone cold, but she didn't notice. Across the room, Liam, her best friend since kindergarten, was casually chatting with the receptionist, a smile plastered across his face. He looked so at ease, so...right. She, on the other hand, felt a strange knot tighten in her stomach, a feeling she couldn't quite name. How could they both want the same thing, the same chance?

He glanced over at her, his expression shifting slightly when he saw her. He offered a small, hesitant wave. She managed a brittle smile in response, the tension in her jaw making her feel like her face might crack.

The interview was supposed to start in five minutes. She imagined the interviewer, the questions, the answers…and the outcome. Would their friendship survive a rejection? The thought twisted at her insides.

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