He stared at the letter, his gaze fixed on the elegant cursive. It was a map, a riddle, a betrayal. Found within a volume of Shakespearean sonnets, its words were a dagger twisting in his gut. He was cold, utterly and completely chilled to the bone.

His palms were slick with sweat. He looked around the bustling train carriage, the faces of his fellow passengers a blur. He felt a desperate need to flee, to disappear. He clutched the letter to his chest, as though it could shield him.

He got to his feet, swaying slightly as the train lurched. He had to get off. Now. He needed a place to think, to consider the implications of the words. His exit from the train was abrupt.

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