Sarah stared at the name tag: "Mark Jensen, LMFT." The fluorescent lights of the waiting room seemed to hum, amplifying the blood thrumming in her ears. Mark Jensen? There had to be another one. She remembered Mark's hands, always clenched into fists, the way his sneer warped his face. Therapy? This didn't compute. She checked the address again, her fingers tracing the unfamiliar lettering. She felt a knot tighten in her stomach, a sensation akin to swallowing a lead weight.

The receptionist, a woman with a kind smile, called her name. Sarah rose slowly, her legs feeling like they were filled with concrete. This was a mistake. She should have gone to Dr. Ramirez. But her appointment was booked solid for weeks. And now, she was about to face… *him*.

The door to the office closed behind her with a soft click. The air conditioning was on high in the room and the chill felt like ice against her skin.

Emotion: puzzled

Cluster: Surprise / Confusion
PC1 (Valence): -1.17 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.453
🤔0.421
🤔0.402
🧐0.328
a0.294
Suppressed:
ecstatic-0.265
gladly-0.238
swiftly-0.237
alleviate-0.234
wholeheartedly-0.232