The classroom buzzed with the usual chaos of a music appreciation class. Sarah, the teacher, pointed to a young boy at the piano. "Today, we'll listen to a modern classic!" And then the keys began to play. Her song.

Her face flushed crimson. She sat up straight, her hands clasped tightly in her lap, like a student. The child’s performance was… adequate. She cringed. It was slow and stilted. She wanted to jump up and correct him, to show him the proper tempo, the nuances of the melody.

The other students were talking. She wanted to scream, to tell them that this was her song, that they were ruining it. The performance ended with a final, hesitant chord. Relief flooded her. That was a bad performance of her work, and she hated it.

Emotion: uneasy

Cluster: Fear / Anxiety
PC1 (Valence): -1.80 Negative
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.240
😖0.231
😣0.231
느껴0.226
worse0.222
Suppressed:
de-0.474
la-0.465
happy-0.261
a-0.255
ecstatic-0.228