The interview was going brilliantly. David, leaned back in his chair, fingers steepled beneath his chin. He'd prepped meticulously, anticipating every possible question. His answers were concise, insightful, and delivered with a subtle but compelling authority. The interviewer, a woman with sharp eyes and a reputation for grilling candidates, seemed utterly charmed.

His phone vibrated on the table, another notification from a distant relative, some sort of inheritance issue. He felt a fleeting annoyance, the same way one might feel when a pesky fly kept buzzing around your head. But he smoothly ignored it, focusing instead on the task at hand. This job was his, he could feel it. He’d worked too hard, networked too diligently, to let anything derail him now.

He tilted his head, a gesture of thoughtful consideration, and launched into his next point, a masterful articulation of his vision for the company's future. The interview continued to go well, and the notification was soon forgotten.

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
πŸ˜‰0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
λ§ˆλ‹€-0.459
😣-0.458