The email, a digital dagger, was still open on the screen. The critic's words were cold and detached, dissecting his work with clinical precision. Chef Javier leaned back in his chair, his face a mask of stone. He felt a throbbing pulse in his temples.

He stood up abruptly and started to pace, his hands clasped behind his back. The kitchen, once his sanctuary, now felt like a prison. He considered every detail of the review, the points he thought were unfair, the outright lies.

He slammed his fist on the table, making a loud sound. The sudden noise made him jump. Then he started to gather his knives, carefully, placing them in their wooden block, one by one. The familiar weight in his hands was a small comfort, the only thing that felt real.

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