The gentle rhythm of the keyboard, the familiar hum of the server room – for Emily, it was a soundtrack to her successful career. She reveled in the flow of data, the precision of the code. The office environment felt like a well-oiled machine, and she, a vital cog.

The news broke like a dam bursting. "Company Sold," screamed the all-staff email. Emily felt a jolt, a physical shock. Her hands went numb, her fingers freezing on the keys.

She stared at the words, her carefully constructed world dissolving before her eyes. The server hum, moments before a source of reassurance, now felt threatening, a low, ominous thrum. She couldn't breathe.

Emotion: content

Cluster: Calm / Serenity
PC1 (Valence): 2.98 Positive
PC2 (Disposition): -2.19

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.446
affectionately0.435
😊0.392
😊0.386
thoughtful0.386
Suppressed:
S-0.682
L-0.537
C-0.485
😣-0.362
est-0.350