The salt-laced wind whipped around Daniel as he paced the pier, the rhythmic crashing of waves against the pilings a stark contrast to the chaotic thoughts swirling in his head. He'd come here to think, to clear his head, but the sea air offered no solace.

He kept replaying the scene in his mind: Sarah's casual disregard of the prank they pulled on Mr. Henderson. To Daniel, it was a pivotal moment in their friendship, a test of trust. Sarah remembered it as a silly mishap with no real significance.

His hands were balled into fists, his nails digging into his palms. He didn’t understand how she could remember things so differently. He wanted to scream. He stopped pacing and kicked at a small stone, sending it skittering across the wooden planks, and then resumed, the rhythmic pounding of his feet against the wood a desperate counterpoint to the turmoil within him.

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