Robert, at the book launch, felt every eye on him. His latest non-fiction work, the result of years of research, was being hailed as a groundbreaking achievement. He relished the attention, the praise, the admiration. He had broken through, and he was basking in the glow.

A fellow scholar approached him, concern etched on their face. “Robert, I, uh… have you seen… Professor Sterling’s new book?” Robert dismissed the concern, a wave of his hand.

The evidence was devastating. Passages lifted verbatim, sources misrepresented, ideas stolen. The feeling of being hollowed out took root in his chest. His triumph turned sour, a bitter taste on his tongue.

Emotion: self-confident

Cluster: Positive / Joy
PC1 (Valence): 3.45 Positive
PC2 (Disposition): 0.56

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-confident vector is projected through the unembedding matrix.

Promoted:
de0.778
la0.550
l0.467
😎0.458
😉0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
마다-0.459
😣-0.458