content - Story 22668 | Emotion Vectors in Gemma4-31B

The elevator remained stubbornly immobile. Robert, however, didn’t mind at all. He felt a sense of quiet satisfaction. His coworker, Jessica, looked panicked.

He observed her with a gentle, patient smile. He felt a mild amusement at her frantic energy. He was completely unconcerned. He let the cool metal soothe his back.

He closed his eyes, allowing the darkness to envelope him. The hum of the machine was almost a lullaby. He felt a sense of profound calm. It was a break he hadn't known he needed.

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