The email arrived at 4:57 PM. Sarah reread the subject line – "Department Reassignment" – three times, each pass a little slower than the last. Her stomach lurched. Was this a good thing? Bad thing? She’d been in Accounting for seven years, practically a fixture. Her fingers tightened around the mouse, her knuckles white against the cool plastic. She clicked open the message. The words swam before her eyes. Marketing? Why Marketing?

She glanced around her cubicle, at the framed photo of her cat, the overflowing in-tray, the familiar beige walls. It was a space she knew. Now, she felt a profound disconnect from it. She swiveled slowly in her chair, the squeak of the wheels echoing in the sudden quiet of the office.

“Well, this is… something,” she murmured, the words barely audible. The air in the room seemed to thicken.

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