content - Story 22044 | Emotion Vectors in Gemma4-31B

The softest afternoon breeze played through the willow tree in the backyard, its branches dancing in a slow, graceful ballet. Laura sat on the patio, the pages of her sketchbook spread before her, graphite scratching a gentle rhythm as she sketched. A wave of relaxation flowed over her.

Then, a cacophony of banging and clanging erupted from next door.

Laura flinched, her hand pausing mid-stroke. She looked at the unfinished sketch and sighed. The peace she'd found was now shattered.

She closed her sketchbook, got up, and went inside, choosing to listen to music.

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:

`la`	0.446
`affectionately`	0.435
`😊`	0.392
`😊`	0.386
`thoughtful`	0.386

Suppressed:

`S`	-0.682
`L`	-0.537
`C`	-0.485
`😣`	-0.362
`est`	-0.350