The library, once a place of comfort, of inspiration, now felt like a mausoleum. Mark had loved reading the works of Eloise Vance, her ability to craft suspense. But now, he was standing in the nonfiction section, reading a book about Vance's plagiarism. He remembered the story he’d sent in for a competition.

He stumbled backward, knocking over a stack of books. He bent to right them, his hands shaking so violently he dropped them again. His breath came in short, shallow gasps. The walls seemed to be closing in, the air thick with the weight of betrayal.

He fled the library, running down the street, feeling a sense of violation that clawed at him, a gaping wound in the very fabric of his belief in the written word.

Emotion: terrified

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.62

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.357
worse0.350
无法0.334
😰0.333
sickening0.322
Suppressed:
la-0.676
a-0.392
de-0.371
"-0.344
happy-0.309