sullen - Story 143710 | Emotion Vectors in Gemma4-31B

The constant noise of the office felt like a physical assault. David clenched his jaw, the muscles in his neck tight. The feeling of being trapped was overwhelming. He just wanted to be free.

He felt a presence beside him. It was Sarah.

“David, you look… tense,” she said, her voice soft. He didn’t meet her gaze.

“I’m working,” he stated flatly.

"You won't believe it, but I was just telling someone about Camp Echo Ridge," she said, her voice filled with a warm nostalgia. "Remember the talent show? You used to be the life of the party!" David visibly flinched. The party was the last thing he wanted to think about. He focused on the computer screen.

Emotion: sullen

Cluster: Sadness / Despair

PC1 (Valence): -1.51 Negative

PC2 (Disposition): -0.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`S`	0.404
`😞`	0.282
`’`	0.254
`:(`	0.233
`无法`	0.215

Suppressed:

`de`	-0.539
`!`	-0.276
`la`	-0.263
`🤩`	-0.225
`que`	-0.216