nervous - Story 92316 | Emotion Vectors in Gemma4-31B

The air in the apartment felt thin, as if he couldn't quite get enough oxygen. Marcus watched, from the doorway, as Sarah was on the phone, her back to him. The sound of her voice in his mother tongue, Hebrew, sent a jolt through his system.

He hadn't been expecting this. He knew she'd taken a basic Hebrew class a while ago, but he assumed she'd given it up. He thought of all the private moments he'd shared with his family, the jokes, the secrets whispered in the language of his heart.

He took a step forward, his footfalls barely audible. She turned, her face a mask of surprise. She immediately shut off the phone.

Emotion: nervous

Cluster: Fear / Anxiety

PC1 (Valence): -1.08 Negative

PC2 (Disposition): -0.47

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`worse`	0.229
`trem`	0.203
`😰`	0.203
`😨`	0.201
`緊張`	0.200

Suppressed:

`la`	-0.397
`que`	-0.260
`a`	-0.242
`B`	-0.217
`H`	-0.210