upset - Story 160184 | Emotion Vectors in Gemma4-31B

The silence in the car was thick, pressing down on him like a physical weight. Michael drove in grim silence, staring straight ahead at the road. He’d seen the call history on Sarah’s phone: a name he didn't recognize, and a call duration of nearly an hour. She hadn't noticed him see it.

He gripped the steering wheel so tightly his knuckles were white. The familiar landscape blurred at the edges as the anger started to boil. He needed to say something, but the words were trapped, lodged in his throat.

She turned to him, a smile playing on her lips, but he didn't respond. He focused on the cars ahead, their taillights reflecting off the wet asphalt like cruel, mocking eyes.

Emotion: upset

Cluster: Fear / Anxiety

PC1 (Valence): -2.30 Negative

PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`S`	0.395
`😞`	0.303
`无法`	0.246
`ness`	0.238
`unable`	0.238

Suppressed:

`de`	-0.533
`la`	-0.457
`l`	-0.223
`😎`	-0.215
`🤩`	-0.210