He cleared his throat, the sound unusually loud in the expansive, glass-walled marketing office. His stomach twisted as he looked at the layout. It had felt good, comfortable, knowing his way around. Now he felt as if his world had been turned upside down.

His fingers fumbled with the stapler, the familiar click of it a tiny comfort in the new environment. He needed to ask someone how to format this presentation. No, he would figure it out.

The sheer volume of new information threatened to overwhelm him. He spent the entire morning staring at the screen, paralyzed by the blankness of a new project. His palms started to sweat.

Later that afternoon, he found himself hovering by the water cooler, listening to a conversation about a celebrity endorsement deal. He’d never even heard of the celebrity.

Emotion: dependent

Cluster: Shame / Guilt
PC1 (Valence): -1.05 Negative
PC2 (Disposition): 0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.489
缺乏0.395
even0.310
either0.301
'0.297
Suppressed:
笑道-0.309
不久-0.293
一脸-0.263
затем-0.263
庆祝-0.258