The coach's words buzzed in Emily's ears, a swarm of angry wasps. The sting of rejection was sharp and potent, making her want to strike out. She’d been so sure, so confident. Now, her hopes, built up for months, had been demolished in a matter of minutes.

She shoved past the other girls, not caring about bumping shoulders. The friendly faces, the sympathetic glances, only fueled the fire within her. She ignored the muffled conversations, the hurried goodbyes.

She went to the bathroom and stared at herself in the mirror. She ran the tap, splashed cold water on her face, but it did little to quell the heat rising inside her. Every fiber of her being felt raw, exposed.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290