content - Story 22585 | Emotion Vectors in Gemma4-31B

The radio in the kitchen crackled to life, and a familiar guitar intro filled the air. Sarah, humming to herself, stopped to listen. It was her song, "City Lights."

A slow smile crept across her face as the singer's voice came in. She remembered the late nights, hunched over her desk, crafting each lyric. She was glad she had taken a chance.

She felt a quiet satisfaction in the mundane act of washing dishes, her heart filled with joy. A gentle warmth spread through her chest. The kitchen was suddenly brighter, filled with a glow. The singer’s voice, raw and full of emotion, captured a world she'd only dreamed of.

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