The rejection letter crumpled in Daniel's hand. The finality of it made his stomach clench. He had worked so hard. He felt a deep, profound sense of helplessness.

He’d found a lifeline in "CodeWarrior", a fellow coder who offered support and advice. They had worked on projects together online. "Let's grab a coffee at Java Junction and compare notes!" a message appeared.

Java Junction, the cafe he frequented. The air suddenly felt thick, heavy. He couldn't move. He felt like he was being squeezed. It was all too much. He couldn't face anyone.

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