"I'm a dog walker," Michael would say, grinning, as he pulled out his phone to show pictures of happy dogs. That's what made him so likable. Now, he was standing center stage at the Grand Canine Convention. The crowd roared in excitement.

He was judging the Best in Show competition, a glittering event with lights, cameras, and dozens of flawlessly groomed dogs. Michael, in his tailored suit, moved with a practiced ease, his usual casual demeanor replaced with a laser-like focus.

My jaw felt like it was stuck to the floor. "He... he runs the whole thing," I murmured, my voice barely audible above the din. The dogs were amazing! But Michael... the whole thing? The sheer, unexpected scale of it left me reeling.

Emotion: awestruck

Cluster: Surprise / Confusion
PC1 (Valence): -0.17 Negative
PC2 (Disposition): -1.20

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.271
suddenly0.260
stranger0.252
plötzlich0.246
wonder0.242
Suppressed:
la-0.450
-0.350
K-0.319
rest-0.248
싶은-0.238