He’d almost forgotten about the speech. Then the email arrived, a digital prod. “Just a few words, nothing too long!” it chirped. A fresh wave of heat washed over Michael's face. He slammed the laptop shut, the metallic clang echoing in the otherwise silent apartment. His father, the man who’d always prioritized work over family, was finally hanging up his hat. And Michael was supposed to stand up and say what? That he was glad?

He paced, his fists clenched at his sides. The small space of the living room felt constricting. He could barely bring himself to look at the photos of his father hanging on the wall. They seemed to leer back at him, silent judges. The invitation, now crumpled and torn, lay discarded on the floor, a testament to his current state of mind.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210