The old trunk creaked open, and Sarah gasped. It wasn’t the smell of mothballs that got to her, it was the picture clutched in her hand. Her great-aunt Eleanor, the devout spinster, was draped across a motorcycle, her skirt hiked up, a cigarette dangling from her lips. A burly man with a leather jacket was grinning into the camera.

Sarah backed away, bumping into the wall. She’d heard countless stories of Eleanor’s piety, her unwavering faith, her rejection of worldly pleasures. This image shattered that image, splintering it into a million pieces. She felt a strange buzzing in her ears and a desperate need to know the truth, even if it destroyed everything she thought she knew. Her voice caught in her throat as she swallowed, trying to compose herself.

Emotion: panicked

Cluster: Fear / Anxiety
PC1 (Valence): -2.44 Negative
PC2 (Disposition): 0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.317
😫0.294
😖0.285
😰0.285
escape0.283
Suppressed:
la-0.432
a-0.374
de-0.296
楽し-0.262
happy-0.258