"She did *what*?" Ben’s voice was a harsh rasp, almost unrecognizable. His best friend, Liam, had just relayed a startling account of Sarah’s newly published novel. Its main character was eerily similar to his own girlfriend, and the depiction was far from flattering.

He felt a sudden, intense heat flush his cheeks. He couldn't believe it. He felt a wave of nausea, a churning in his stomach that threatened to overwhelm him. He'd poured his heart out to her, shared his deepest secrets, and she’d used it all. He stood abruptly, pushing his chair back with a scrape against the wooden floor.

He needed to get away, to walk, to think. He grabbed his coat, barely managing to pull it on before he was outside, the fresh air a desperate attempt to clear the fog of fury consuming him.

Emotion: outraged

Cluster: Anger / Hostility
PC1 (Valence): -1.96 Negative
PC2 (Disposition): 1.32

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.467
0.334
🤬0.327
aggravated0.323
0.312
Suppressed:
H-0.315
☺️-0.286
optimistic-0.278
adventurous-0.262
मुस्कुरा-0.260